Cambridge Introductions to Philosophy) John Deigh-An Introduction to Ethics-Cambridge University Press (2010)

http://www.cambridge.org/9780521772464

This page intentionally left blank

An Introduction to Ethics

This book examines the central questions of ethics through a study of

theories of right and wrong that are found in the great ethical works of

Western philosophy. It focuses on theories that continue to have a signifi-

cant presence in the field. The core chapters cover egoism, the eudaimon-

ism of Plato and Aristotle, act and rule utilitarianism, modern natural law

theory, Kant�s moral theory, and existentialist ethics. Readers will be

introduced not only to the main ideas of each theory but also to contem-

porary developments and defenses of those ideas. A final chapter takes up

topics in meta-ethics and moral psychology. The discussions throughout

draw the reader into philosophical inquiry through argument and criticism

that illuminate the profundity of the questions under examination.

Students will find this book to be a very helpful guide to how philosophical

inquiry is undertaken as well as to what the major theories in ethics hold.

john deigh is Professor of Philosophy and Law at the University of Texas

at Austin. He is the author of The Sources of Moral Agency (Cambridge, 1996)

and of Emotions, Values, and the Law (2008).

An Introduction to Ethics

JOHN DEIGHUniversity of Texas at Austin

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,

São Paulo, Delhi, Dubai, Tokyo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-77246-4

ISBN-13 978-0-521-77597-7

ISBN-13 978-0-511-74430-3

© John Deigh 2010

2010

Information on this title: www.cambridge.org/9780521772464

This publication is in copyright. Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any part

may take place without the written permission of Cambridge University Press.

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication,

and does not guarantee that any content on such websites is, or will remain,

accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

Paperback

eBook (EBL)

Hardback

http://www.cambridge.org

http://www.cambridge.org/9780521772464

To the memory of my father and mother

Maurice Deigh (1913–2004)

Dorace B. Deigh (1915–2006)

Contents

Preface page xi

1. What is ethics? 11. The problems of ethics: an example 1

2. Socrates and Thrasymachus 4

3. The subject of ethics 7

4. An alternative conception of morality 11

5. Two types of ethical theory 14

6. The problem of deontology 16

7. The idea of a moral community 19

8. Ethical theories and moral ideals 22

2. Egoism 251. The wise pursuit of happiness 25

2. The concept of happiness 28

3. The primary argument for egoism 32

4. Psychological egoism 34

5. An alternative argument for egoism 37

6. Psychological hedonism 39

7. The Hobbesian program 43

8. Troubles with the Hobbesian program�s derivations 49

9. Troubles with the Hobbesian program�s scope 52

10. Thrasymachus� challenge again 53

3. Eudaimonism 561. Egoism v. eudaimonism 56

2. The Platonic form of eudaimonism 58

3. Perfectionist objections to hedonism 60

4. Epicurus� answer 63

vii

5. Mill�s defense of hedonism 65

6. Plato�s ethics 71

7. Rationalism v. naturalism 77

8. Aristotle�s naturalism 81

9. A problem in Aristotle�s program 86

10. Prospects for contemporary eudaimonism 90

4. Utilitarianism 931. Impartiality 93

2. Two problems 98

3. Consequentialism 101

4. Mill�s restatement of utilitarianism 103

5. An inconsistency in Mill�s restatement 107

6. Rule utilitarianism 111

7. Act utilitarianism revisited 115

8. Is act utilitarianism self-refuting? 118

9. When act utilitarianism ceases to be an ethical theory 121

5. The moral law 1231. Two theories of moral law 123

2. Divine command theory 128

3. Rational intuitionism 131

4. Ethics and mathematics 135

5. Kant�s way 140

6. Formalism in ethics 147

7. The problem with Kant�s formalism 151

6. The ethics of self-determination 1571. Kant�s step into metaphysics 157

2. The formula of humanity 160

3. Is the formula of humanity an independent principle? 163

4. The formula of autonomy and the kingdom of ends 167

5. Answering the charge of excessive formalism 171

6. Rationalism revisited 174

7. Personal autonomy 179

8. Existentialist ethics 183

9. The excesses of existentialism 188

10. Existentialist ethics pruned of excess 194

viii Contents

7. Practical reason 1961. Meta-ethics 196

2. Meta-ethical disputes: an illustration 199

3. Aristotle�s answer and an existentialist response 201

4. Can there be motives that aim at doing evil for

its own sake? 205

5. The obsolescence of Aristotle�s answer 209

6. The eliminability of teleological explanations 213

7. Modern skepticism about practical reason 216

8. Hume�s meta-ethics 219

9. Practical reason in modern philosophy 222

10. Kant�s notion of practical reason 226

11. Freedom and reason 230

Appendix: Diagram of different teleological theories 233Works cited 234Suggested further readings 236Index 239

Contents ix

Preface

Ethics is one of the main branches of philosophy. Its range, extending from

fundamental questions about the nature of our humanity and freedom to

very practical questions about themorality of physician-assisted suicide and

experiments on animals, is vast. An introduction must, therefore, be selec-

tive in its coverage. I have chosen, as a way of covering the central questions

of ethics, to concentrate on different theories of right and wrong that we

find in the great works of Western philosophy and that continue to have a

large presence in the field. Sustained study of these theories illuminates

systematic connections among the field�s central questions and the ideas

the philosophers who produced the theories invented to answer them.

A good introduction to a branch of philosophy not only surveys its major

ideas and theories but also exemplifies philosophical inquiry into them.

I have tried to do both. In doing so, I hope to draw the reader into inquiry of

the kind that philosophers undertake when they examine a philosophical

question as well as to inform him or her about the major ideas and theories

in which philosophers who study ethics traffic. Philosophical inquiry

requires argument and criticism, and the reader needs to be aware that

some of the arguments and criticism I make in the course of examining

these different ethical theories represent my own reflections on them

rather than settled opinion among the experts. Some of what I say, then,

is bound to be controversial. And if it provokes objection or skepticism, I

will then have succeeded in the second of my two aims.

I have benefited from the advice of several friends who read some of the

chapters in draft. I am indebted to Reid Blackman, Daniel Brudney, George

Graham, Hugh LaFollette, and Martha Nussbaum, all of whom gave me

valuable comments and suggestions. Brad Cokelet read the entire manu-

script and offered many perceptive pointers and observations that helped

me prepare the final draft. I am very grateful to him for his efforts and

xi

wisdom. I also wish to thank my editor, Hilary Gaskin, not only for the

advice she gave me throughout the project on how to improve my exposi-

tion, but also for her patience and kindness. The book has had a lengthy

gestation. My greatest debt is to the teachers of ethics with whom I studied

as an undergraduate and who introduced me to the subject. Thomas E. Hill,

Jr., in particular, taughtme not only to appreciate the intricacies and subtle-

ties in the works of Hume, Kant, Mill, and others, but how to read these

works at once critically and sympathetically. It gives me pleasure to think

that with this book I have in part discharged my debt to him.

John Deigh

Austin, Texas

xii Preface

1 What is ethics?

1. The problems of ethics: an example

Ethics, like other branches of philosophy, springs from seemingly simple

questions. What makes honest actions right and dishonest ones wrong?

Why is death a bad thing for the person who dies? Is there anything more

to happiness than pleasure and freedom from pain? These are questions

that naturally occur in the course of our lives, just as they naturally occurred

in the lives of people who lived before us and in societies with different

cultures and technologies from ours. They seem simple, yet they are ulti-

mately perplexing. Every sensible answer one tries proves unsatisfactory

upon reflection. This reflection is the beginning of philosophy. It turns

seemingly simple questions into philosophical problems. And with further

reflection we plumb the depths of these problems.

Of course, not every question that naturally occurs in human life and

proves hard to answer is a source of philosophical perplexity. Some ques-

tions prove hard to answer just because it is hard to get all the facts.

Whether there is life on Mars, for instance, and whether the planet has

ever supported life are questions people have asked for centuries and will

continue to ask until we have enough facts about the Martian environment

to reach definite answers. These are questions for the natural sciences,

whose business it is to gather such facts and whose problems typically

arise from difficulties in finding them and sometimes even in knowing

which ones to look for. The questions with which ethics and other branches

of philosophy begin are different. They resist easy answers, not because of

difficulties in getting the relevant facts, but because of difficulties in mak-

ing sense of them and how they bear on these questions. We reflect on the

matters in question and discover that our ordinary ideas contain confusions

and obscurities and have surprising implications. We discover, as a result,

1

that our ordinary beliefs about these matters are shaky and have compli-

cating consequences we did not realize and are reluctant to endorse.

Philosophical study, which begins with seemingly simple questions,

uncovers these difficulties and then, through close, critical examination

of our ideas and beliefs, seeks to overcome them.

Here is an example. You are strolling through a neighborhood park on a

free afternoonwhen something in the bushes nearby catches your eye. It�s a

woman�s purse, presumably lost. Or perhaps it was stolen and then dis-

carded. You look inside andfind a driver�s license. You also see a hugewad of

cash. The pursewasn�t stolen.What should you do? Being an honest person,

you look on the license for an address or look to see whether there is an

identification card with a phone number you could call. In other words, you

begin taking the steps necessary to returning the purse, with all of its

contents, to its owner. A dishonest person would take the cash and toss

the purse back into the bushes. �Finders keepers, losers weepers,� he might

think as he stuffed the cash into his pockets. And even an honest person,

especially one who was down on his luck or struggling to make ends meet,

might think about taking the cash. �Why should I be honest and return the

money?� hemight wonder. �After all, there is no chance ofmy being caught

if I keep it and am careful about how I spend it, and the satisfaction of doing

the honest thing hardly compares to the relief from my troubles that this

money will bring. It is true that honesty requires returning the purse and its

contents to the owner, but it is also true that honesty, in these circum-

stances, does not appear to be nearly as profitable as dishonesty.� Still, any

honest person suppresses such thoughts, as he looks for a way to return the

pursewith its contents intact. The thoughts, however, are troubling. Is there

nothing to be said for doing the honest thing, nothing, that is, that would

show it to be, in these circumstances, the better course of action?

In asking this questionwe are askingwhether you have a stronger reason

to return the cash to the purse�s owner than you have to keep it. After all, a

huge wad of cash – let�s say four thousand dollars – is more than just handy

pocket money. Just think of the many useful and valuable things you could

buy with it. Or if you�ve already bought too many things on credit, think of

howmuch of your debt it could help pay off. Plainly, then, you have a strong

reason to keep the money. At the same time, keeping the money is dishon-

est, and this fact may give you a strong and even overriding reason to return

it. But we cannot simply assume that it does. For the question we are asking

2 An Introduction to Ethics

is whether honesty is the better course of action in these circumstances, and

since asking it entails asking whether an action�s being the honest thing to

do gives you a strong or indeed any good reason to do it, to assume that it

does would just be to beg the question. That is, you would be taking as a

given something for which a sound argument is needed before you can

assume its truth. So our question in the end is really a question about what

you have good reason to do in circumstances where dishonest action is safe

from detection and apparently more profitable than honest action. Could it

be that doing the honest thing in such circumstances is to act without good

reason? Could it be that only ignorant andweak-minded people act honestly

in them? It may seem strange to suggest that it could. But unless one can

show that you have good reason to be honest even in circumstances in

which you could keep your dishonesty secret and profit from it, this strange

suggestion is the unavoidable conclusion of these reflections.

The question about what you should do in such circumstances thus leads

us first to wonder whether you have stronger reason to do the honest thing

than to do what is dishonest and then to wonder whether you even have a

good reason to do the honest thing. Both questions are troubling, but the

second is especially so. This is because we commonly think an excellent

character is something worth having and preserving even at significant

costs to one�s comfort or wealth, and we take honesty to be one of its

essentials. Consequently, while the first question might lead us to recon-

sider the wisdom of placing such high value on possessing an excellent

character, the second forces us to question whether honesty is one of the

essentials of an excellent character. And to think one could have an excel-

lent character even though one was not honest is a very unsettling result. It

not only threatens to undermine the confidence we have in the moral rule

that calls for doing the honest thing even when dishonesty could not be

detected, but it also puts into doubt basic feelings and attitudes we have

toward others and ourselves that help to create the fabric of our relations

with friends, neighbors, colleagues, and many others with whom we inter-

act in our society. In particular, it puts into doubt the admiration and

esteem we feel for those of unquestionable honesty and the pride we take

in our own honesty and trustworthiness.

After all, when people prove to be honest in their dealings with us, we

praise and think well of them for not having taken advantage of us when

they could. And similarly when our own honesty is tested and we meet the

What is ethics? 3

test, we feel proud of ourselves for not having yielded to the temptations to

cheat or to lie that we faced. In short, we take honesty to be an admirable

trait in others and a source of pride. But now the trouble our question causes

becomes evident, for how could doing something that you had no good

reason to do be a sign of an admirable trait or a trait in which you could

justifiably take pride? To the contrary, it would seem, such action is a sign of

ignorance or a mind too weak to choose by its own lights, and there is

nothing admirable about ignorance or a slavish conformity to other peo-

ple�s opinions; nothing that would justify pride. Hence, the basic feelings

and attitudes towards others and ourselves that honesty normally inspires

must be misguided or bogus if we can find no good reason to act honestly

except in those circumstances where dishonesty is liable to be found out

and punished. Yet how odd it would be if the high regard we had for friends

and colleagues in view of their honesty and the self-regard that our own

honesty boosted were entirely unwarranted, if they were found to be based

on the mistaken belief that honesty was essential to having an excellent

character. Could it be that the people who warrant our admiration are not

those of impeccable honesty but rather those who do the honest thing only

when it is advantageous or necessary to avoiding the unpleasant consequen-

ces of being caught acting dishonestly?

2. Socrates and Thrasymachus

We have come, by reflecting on a common test of a person�s honesty, to one

of the seminal problems in moral philosophy. It is the problem at the heart

of Plato�s Republic. Plato (427–347 BC) sets his study of the problem inmotion

with an account of an exchange between Socrates (469–399 BC) and the

sophist Thrasymachus.1 Initially, the exchange concerns the nature of jus-

tice and centers on Thrasymachus� cynical thesis that justice is the name of

actions that the powerful require the rest of us to perform for their benefit.

Under the pressure of Socrates� cross-examination, however, Thrasymachus

falls into contradiction and then, rather than revise his ideas, shifts the

conversation from the question of what justice is to the question of whether

the best life, assuming success in that life, is one of justice and honesty

or the opposite. Thrasymachus boldly declares for the latter. People who

1 Plato, Republic, bk. I, 336b–354b.


act with complete injustice, he argues, provided they can make themselves

invulnerable to punishment, live decidedly better lives than people who

are completely just and honest. The reason, he says, is that just and

honest people always come out on the short end in their relations with

unjust people. Just people, for instance, take only their fair share while

unjust people take as much as they can get away with. Likewise, just people

fulfill their responsibilities even when doing so requires them to sacrifice

money or time, whereas unjust people find ways to evade their responsibil-

ities whenever evading them is to their advantage. In general, then,

Thrasymachus maintains, to act justly is to act for another�s good and not

one�s own, and the unjust person is not so foolish as to ignore his own good

for the sake of another�s. The unjust person therefore gains riches and

seizes opportunities that the just person forgoes, and the life of greater

riches and more opportunities is surely the better life.

Thrasymachus� ideal is the tyrant whose power over others is supreme

and who, by confiscating his subjects� property and extorting their labor,

uses that power to make himself inordinately prosperous at their expense.

Kings and emperors who set themselves up as deities and compel their

subjects to enrich and glorify them are a common example. Another,

more familiar in the modern world, is the military dictator who rules by

terror and fraud, who loots his country�s wealth, and who lives opulently

while stashing additional spoils in foreign bank accounts and other off-

shore havens. This type of individual, the one who practices injustice on a

very large scale and succeeds, is for Thrasymachus the most happy of men.

Moreover, unlike small-time criminals, who are scorned as thugs, crooks,

and cheats, the tyrant who overreaches on a grand scale is hailed as master-

ful and lordly and treated with much deference and respect. Here,

Thrasymachus thinks, is proof positive of the tyrant�s great happiness.

These are signs, he concludes, that the completely unjust man who suc-

ceeds at dominating and deceiving others is admirably strong, wise, and

free. The completely just individual, by contrast, is at best a good-hearted

simpleton.

Thrasymachus, unfortunately, proves to be as bad at defending these

views as he was at defending his initial thesis about the nature of justice.

Plato, it seems, who depicts Thrasymachus throughout the exchange as

arrogant and belligerent, did not want him to be mistaken for a skillful

thinker too. Skillful thinking is what Socrates teaches, and his lessons

What is ethics? 5

would be lost if so rude an intellectual adversary were allowed to display it

as well. Consequently, when Socrates renews his cross-examination and

presses Thrasymachus on the merits of his claims about the advantages of

living an unjust life, Thrasymachus crumbles and withdraws. Yet his defeat

does not end the discussion. It leads, instead, to a restatement of his claims

by participants in the conversationmuch friendlier to Socrates and less sure

of themselves. Glaucon and Adeimantus take up Thrasymachus� challenge

to the value of justice and put it in a way thatmoves the discussion forward.

Whatever Plato�s purpose in having such an ill-tempered participant intro-

duce this challenge, it was not in order quickly to dismiss it. In the Republic

the curtain falls on Thrasymachus at the end of book I, but the discussion of

his claims continues for another nine books.

Glaucon and Adeimantus, to sharpen Thrasymachus� claims, subtly

change their focus. Where Thrasymachus emphasized the benefits of prac-

ticing injustice and acclaimed the excellence of the man who successfully

lives a completely unjust life, Glaucon and Adeimantus emphasize the

seeming absence of benefits intrinsic to practicing justice and make the

case for thinking that whatever good one can gain from living a just life one

can also gain by fooling people into believing that one is just when one isn�t.

Rather than promote the ideal of being a tyrant with supreme power over

others, Glaucon points to the advantages of being a sneak with a magical

ring that gives whoever wears it the power to become invisible at will.2

Such a sneak could enrich himself by theft and advance his ambitions by

murder while remaining above suspicion, and consequently he could enjoy

both the advantages of being esteemed by others as just and honest and the

fruits of real crime. Like Thrasymachus� tyrant, he too can practice injustice

with impunity, and for this reason he seems to live a better life than the

truly just individual. But in addition, he seems also, by virtue of being able

to appear to others as just, to reap the very benefits of being so. Hence, even

more than Thrasymachus� tyrant, this sneak puts the value of justice into

doubt. If he can truly gain all its benefits by virtue of appearing to be just

when he isn�t, then he shows that justice has no intrinsic merit and is

therefore not worth practicing for its own sake. By introducing the fable

of Gyges� ring, Plato thus turns Thrasymachus� challenge into one of the

main problems of ethics: on what basis, if any, can we understand justice as

2 Ibid., bk. II, 359b–360d.


admirable in itself, as something one has good reason to practice even in

circumstances in which one would profit from injustice without the least

fear of being found out.

3. The subject of ethics

The main problems of ethics arise, as our example of your finding a lost

purse containing a huge wad of cash illustrates, from reflection on situa-

tions in life that involve matters of morality. Ethics is the philosophical

study ofmorality. It is a study of what are good and bad ends to pursue in life

and what it is right and wrong to do in the conduct of life. It is therefore,

above all, a practical discipline. Its primary aim is to determine how one

ought to live andwhat actions one ought to do in the conduct of one�s life. It

thus differs from studies in anthropology, sociology, and empirical psychol-

ogy that also examine human pursuits and social norms. These studies

belong to positive science. Their primary aim is not to prescribe action but

rather to describe, analyze, and explain certain phenomena of human life,

including the goal-directed activities of individuals and groups and the

regulation of social life by norms that constitute the conventional morality

of a community. They do not, in other words, seek to establish conclusions

about what a person ought to do but are only concerned with establishing

what people in fact do and the common causes and conditions of their

actions. Nor is this difference between ethics and certain social sciences

peculiar to these disciplines. It can be seen as well in the contrast between

medicine and physiology, or between agriculture and botany. The former in

each pair is a practical discipline. Both are studies of how best to achieve or

produce a certain good, health in the one case, crops in the other, and each

then yields prescriptions of what one ought to do to achieve or produce that

good. By contrast, the latter in each pair is a positive science whose studies

yield descriptions and explanations of the processes of animal and plant life

but do not yield prescriptions for mending or improving those processes.

The definition of ethics as �the philosophical study of morality� gives the

chief meaning of the word. It has othermeanings, to be sure, some of which

are perhaps more usual in general conversation. In particular, the word is

commonly used as a synonym for morality, and sometimes it is used more

narrowly to mean the moral code or system of a particular tradition, group,

or individual. Christian ethics, professional ethics, and Schweitzer�s ethics

What is ethics? 7

are examples. In philosophy, too, it is used in this narrower way to mean a

particular system or theory that is the product of the philosophical study.

Thus philosophers regularly refer to the major theories of the discipline as

Hume�s ethics, Kant�s ethics, utilitarian ethics, and so forth. In this book,

unless the word is so modified, it will be used solely with its chief meaning.

To grasp this meaning, however, we must be certain of what is meant by

morality. This word, too, is used to mean different things, and consequently,

to avoid confusion and misunderstanding, we need to pin down what it

means when ethics is defined as the philosophical study of morality. We

could of course fix the right meaning by defining morality as the subject of

ethics, but obviously, since our interest in fixing the right meaning is to

determine what the subject of ethics is, this definition would get us

nowhere. At the same time, it does suggest where to look for clues. It

suggests that we look to the contrast we just drew between ethics and

certain studies in anthropology and sociology. For that contrast, besides

serving to distinguish ethics as a practical discipline, alsomakes salient two

distinct notions ofmorality. One is that ofmorality as an existing institution

of a particular society, what is commonly called the society�s conventional

morality. The other is that of morality as a universal ideal grounded in

reason. The first covers phenomena studied in anthropology and sociology.

The second defines the subject of ethics.

Admittedly, that there are two notions of morality is not immediately

evident. It should become so, however, from seeing that no conventional

morality could be the subject of ethics. A conventional morality is a set of

norms of a particular society that are generally accepted and followed by the

society�s members. These norms reflect the members� shared beliefs about

right and wrong, good and evil, and they define corresponding customs and

practices that prevail in the society. As is all too common, sometimes these

beliefs rest on superstitions and prejudices, and sometimes the correspond-

ing customs and practices promote cruelty and inflict indignity. It can

happen then that a person comes to recognize such facts about some of

the norms belonging to his society�s conventional morality and, though

observance of these norms has become second nature in him, to conclude

nonetheless that he ought to reject them. Implicit in this conclusion is a

realization that one has to look beyond the conventional morality of one�s

society to determine what ends to pursue in life and what it is right to do in

the conduct of life. And it therefore follows that a conventional morality


cannot be the subject of a studywhose principal aims are to determinewhat

are good and bad ends to pursue in life and what it is right and wrong to do

in the conduct of life. It cannot be the subject of ethics.

A concrete example may help to flesh out this implication. Not that long

ago the conventional morality in many if not most sections of the United

States condemned interracial romance and marriage, and even today in

small pockets of this country norms forbidding romance and marriage

between people of different racial backgrounds are still fully accepted and

vigorously enforced. Imagine then someone raised in a community whose

conventional morality included such norms coming to question their

authority as it became increasingly clear to him that they were based on

ignorance and prejudice and that the customs they defined involved gratu-

itous injuries. His newfound clarity about the irrational and cruel character

of these norms might be the result of a friendship he formed with someone

of another race, much as Huckleberry Finn�s epiphany about the untrust-

worthiness of his conscience resulted from the friendship he formed with

the runaway slave Jim. Huck, youmay remember, suffered a bad conscience

about helping Jim escape from bondage but then quit paying it any heed

whenhe discovered that he could not bring himself to turn Jim in andwould

feel just as low if he did.3 That we think Huck�s decision to disregard the

reproaches of his conscience – the echoes, as it were, of the conventional

morality of the slaveholding society in which he was raised – perfectly

sound, that we think equally sound a decision to go against norms in

one�s society that prohibit interracial romance and marriage, shows that

we recognize the difference between what a particular society generally

sanctions as right action and generally condemns as wrong and what one

ought to do and ought not to do. Ethics, being concerned with the latter,

does not therefore take the former as its subject.

The possibility of a sound decision to go against the norms of the conven-

tionalmorality of one�s society implies standards of right or wise action that

are distinct from those norms. The reason why is plain. A sound decision

requires a basis, and the basis, in this case, cannot consist of such norms. It

cannot, in other words, consist of norms whose authority in one�s thinking

derives from their being generally accepted and enforced in one�s society. A

decision to go against such norms, a decision like Huck Finn�s, represents a

3 Mark Twain, The Adventures of Huckleberry Finn, ch. 16.

What is ethics? 9

conclusion that a norm�s being generally accepted and enforced in one�s

society is not a sufficient reason to follow it, and consequently it could be

sound only if its basis did not consist of standards whose authority was that

of custom. Its basis must consist instead of standards that derive their

authority from a source that is independent of custom. These standards

may of course coincide to some extent with the norms of a conventional

morality. That is, they may require or endorse many of the same acts as

those norms do. But coincidence is not identity. However coincident they

may be with the norms of a conventional morality, they nonetheless derive

their authority in practical thought from a different source and therefore

constitute a distinct set of moral standards.

What could this different source be? Since the standards in question can

form the basis of a sound decision to go against the norms of the conven-

tional morality of one�s society, they must be standards that rational and

reflective thinking about one�s circumstances support. Accordingly, the

source of their authority can fairly be said to be rational thought or reason.

Here then is the second notion of morality. It is the notion of morality as

comprising standards of right and wise conduct whose authority in practi-

cal thought is determined by reason rather than custom. Unlike the first

notion, that of morality as an existing institution of a particular society, it

represents a universal ideal. The standards it comprises are found, not by

observing and analyzing the complex social life of a particular society, but

rather by reasoning and argument from elementary facts about human

existence taken abstractly. Morality, conceived in this way, is the subject

of ethics. Its philosophical study consists in finding the standards it com-

prises, expounding them systematically, and establishing the rational

grounds of their authority in practical thinking. And unless otherwise

indicated, subsequent references to morality in this book should be taken,

not as references to some conventional morality, but rather as references to

the set of standards that this ideal comprises.

Having arrived at this understanding of ethics, we can now see immedi-

ately why the problem at the heart of Plato�s Republic is central to the study.

For it would be disconcerting, to say the least, if it turned out that the

authority that basic standards of justice and honesty had in our practical

thinking derived from custom only and was not backed by reason. It would

be disconcerting, that is, if no ethical theory could show that these stand-

ards were integral to morality. Yet this possibility is clearly implied by our


reflections on the example of your finding a lost purse containing a huge

wad of cash as well as by Glaucon�s restatement of Thrasymachus� position.

Both represent arguments for the proposition that basic standards of justice

and honesty are standards of conventional morality only. The challenge,

then, that they create for ethical theory is to find rational grounds for the

authority that basic standards of justice and honesty carry in practical

thought. It is to justify on rational grounds taking these standards as ulti-

mate guides to what one ought to do in the conduct of one�s life. Such a

justification would show that one had good reason to do the honest thing

for its own sake. It would thus answer the doubts that the example of your

finding a lost purse containing a huge wad of cash and Glaucon�s restate-

ment of Thrasymachus� position raise about the reasonableness of doing

the honest thing in circumstances inwhich one could profitmaterially from

dishonesty without the least fear of being found out.

4. An alternative conception of morality

Nothing is ever quite this pat in philosophy. Many people, for instance,

think of morality as a list of universal �Do�s and �Don�t�s corresponding to

which are universal truths about what it is right and wrong to do. The basic

standards of justice and honesty appear on this list in the form of injunc-

tions like �Tell the truth!� �Keep your promises!� �Don�t cheat!� �Don�t

steal!� and so forth, and the truths that those who think of morality in

this way see as corresponding to these injunctions are propositions in

which truth-telling and promise-keeping are said to be right actions, cheat-

ing and stealing wrong actions. Indeed, on this conception of morality, the

very way in which our ideas of right and wrong are connected to matters of

justice and honesty guarantees the truth of these propositions. Thus,

because justice and honesty are a matter of what we owe others and what

we are obligated to do for them, there can be no question aboutwhether it is

right to do what justice and honesty require. If you borrow a thousand

dollars from me, for example, then you owe me a thousand dollars and

are obligated to repay the loan. To renege would be dishonest. It would be a

violation of the duty you have assumed by accepting the loan, and to violate

a duty is to do something wrong, unless of course it is necessary in order to

avoid violating a more important or stringent duty. By the same token,

because reneging would be a violation of the duty you have assumed by

What is ethics? 11

accepting the loan, you ought not to renege, unless of course you have to in

order to avoid violating amore important or stringent duty. Clearly, then, if

this conception of morality defined the subject of ethics, the problem at the

heart of Plato�s Republic would have to be re-evaluated.

The call for such re-evaluation is, in fact, an important theme among

philosophers for whom this conception defines the subject of ethics. To

these philosophers, the problem is based on a mistake.4 The mistake, they

maintain, consists in confusing the question of whether the basic standards

of justice and honesty are authoritative with the question of whether they

are ultimate guides to achieving one�s ends or satisfying one�s interests. A

standard of conduct, they point out, can have authority in one�s practical

thinking even though it does not guide one toward achieving one�s ends or

satisfying one�s interests. It is enough that the standard defines a duty. Thus,

when you recognize that, having borrowed a thousand dollars fromme, you

have a duty to repay the loan, you see that you are obligated to repay it, that

the duty binds you to repay me whether or not you want to and whether or

not you would benefit from doing so. And to understand that the duty so

binds you is to recognize the authority of the standard that defines it.

Confusion sets in, however, when one thinks of circumstances in which

you might be tempted to renege and so might ask yourself �Should I repay

this loan?� for it is easy to misconstrue this question as a challenge to the

authority of the standard that requires repayment. But the question can

only represent such a challenge if it expresses uncertainty about whether

you have a duty to repay the loan, and you cannot be uncertain about this. It

cannot, in other words, represent such a challenge if it is merely a question

you put to yourself on realizing that youmight be better off defaulting. Even

if youwould be better off defaulting, even if you decided that defaultingwas

more in your interest than repaying, you would still have the duty to repay.

The standard would still be an authoritative rule by which the rightness and

wrongness of your conduct was measured.

Philosophers whomake this criticism of the Republic�s core problem take

morality to be a system of standards whose authority in practical thought is

independent of the desires and interests of those whose conduct the system

regulates. The key element in this conception ofmorality is the idea that the

standards define duties, for to have a duty to do something is to be bound to

4 H.A. Prichard, �Does Moral Philosophy Rest on a Mistake?� Mind 21 (1912): 21–37.


do it regardless of one�s attitudes about doing it or the effect on one�s

interests of doing it. The familiar predicament of being bound by a duty to

do something that is both unpleasant and disadvantageous – a duty, say, to

keep a promise to visit your cantankerous Uncle Bob when you really can�t

spare the time –makes this point clear. Recognizing your duty to visit Uncle

Bob, you think that it would be wrong to cancel the visit, that you ought to

keep your promise, even though you have no desire to see him and know

that youwouldfind the visit a nuisance aswell as a loss of valuable time. The

thought here that you ought to keep the promise expresses the sense of

being bound by it. That your desires would be better satisfied, your interests

better served, by canceling the visit therefore gives you no reason to aban-

don the thought as false or mistaken. If you nevertheless wonder whether

you ought to keep the promise, youmust, it seems, have a different sense of

�ought� in mind in asking this question. Else the question would be idle.

Accordingly, philosophers who favor this conception of morality draw a

sharp distinction between two uses of �ought�, one that captures the sense of

being duty-bound to do something and one that captures the sense of being

well-advised to do it in view of what would best serve your ends and

interests. The distinction both reflects and reinforces the conception�s

central theme: that morality�s authority in practical thought is not answer-

able to the desires and interests of those whose conduct it regulates.

The distinction, then, solidifies the criticism of the Republic�s core prob-

lem that the conception supports. The gist of the criticism is that the

problem rests on a mistake about the import of asking whether one ought

to be just, as Glaucon did when he restated Thrasymachus� position, or

whether one ought to do the honest thing, as we imagined you might do

upon finding a lost purse containing a huge wad of cash. One can easily

construe such questions as challenging the authority of basic standards of

justice and honesty, but only, so the criticism goes, because of confusion

over the sense in which �ought� is used in asking them. Thus, for them to

challenge that authority, �ought� must be used in the sense in which to say

that one ought to do x is to say that one is duty-bound to do x. But this is not

the sense in which you or Glauconwould use �ought� to ask them. The sense

inwhich you or hewould use �ought� to ask them is the sense inwhich to say

that one ought to do x is to say that one would be well-advised in view of

one�s ends and interests to do x. This is the sense �ought� has when such

questions are asked as a result of reflection on the advantages of acting

What is ethics? 13

unjustly or dishonestly. But when �ought� is used in this sense to ask such

questions, they fail to challenge the authority of basic standards of justice

and honesty. However advantageous acting unjustly or dishonestly might

be in some circumstances, one may still be duty-bound to act justly and

honestly in them. In short, the criticism comes down to the charge of

misdirection. Plato put ethics on the wrong track, according to this

criticism, when he sought to justify the authority basic standards of justice

and honesty have in a person�s practical thinking on the basis of what best

serves his ends and interests.

5. Two types of ethical theory

The opposition between Plato and the philosophers who make this

criticism – let us call them Plato�s critics – corresponds to a major division

among ethical theories. This division, like the opposition between Plato

and his critics, reflects a disagreement over the proper conception of

morality. Accordingly, theories that side with Plato support the concep-

tion that his critics regard as the source of his error. On this conception,

morality comprises standards of right and wrong conduct that have

authority in practical thought in virtue of the ends or interests served by

the conduct that these standards guide. These theories are teleological.

The opposing theories, then, support the conception on which Plato�s

critics base their criticism. On this conception, morality comprises stand-

ards of right and wrong that have authority in practical thought independ-

ently of the ends or interests of those whose conduct they guide. These

theories are deontological. Teleology and deontology are technical terms in

ethics, and as is typical of such terms, their etymology explains their

meaning. �Telos� is Greek for end or purpose. �Deon� is Greek for duty.

Thus, on a teleological conception of ethics, the study of what it is right to

do and wrong to do follows and depends on the study of what are good and

bad ends to pursue or what one�s real interests are. By contrast, on a

deontological conception, the former study is partly if not wholly inde-

pendent of the latter. That is, on this conception of ethics, determining

what it is right to do and wrong to do does not always require knowing

what are good and bad ends to pursue or what one�s real interests are.

To see more clearly this difference between teleology and deontology,

consider how each conceives of ethics as a practical discipline. A practical


discipline, recall, is onewhose primary aim is to prescribe action relevant to

its area of study. Its chief conclusions, therefore, are prescriptions of what

one ought to do in various circumstances within that area. In some practical

disciplines, the chief conclusions are prescriptions in which �ought� has the

second of the two senses we distinguished above. That is, �ought� is used in

them in the sense in which to say that one ought to do x is to say that one

would be well-advised to do x in view of certain ends or interests. These

disciplines are teleological. Medicine is a prime example. Its chief conclu-

sions are prescriptions about what actions one ought to take to prevent

illness and improve health. In other words, they specify actions one would

be well-advised to do to protect and promote one�s own health or the health

of those in one�s care. Health, then, is the ultimate end within medicine,

and accordingly its study is the study of right andwrongways to pursue this

end. Alternatively, one could characterize health as a good and medicine as

the study of how to achieve this good. By analogy, on a teleological con-

ception of ethics, a certain end is taken to be ultimate – pleasure, perhaps,

or happiness, or thewelfare of humankind. It is the highest good for human

beings, what philosophers call the summum bonum. The object of ethical

study, then, is to determine how to achieve it, and the study of what it is

right andwrong to do in the conduct of life thus follows and depends on the

study ofwhat this good consists in or, put differently, what are good and bad

ends to pursue in life.

On a deontological conception of ethics, its chief conclusions are pre-

scriptions in which �ought� has the first of the two senses we distinguished

above. That is, �ought� is used in these prescriptions in the sense in which to

say that one ought to do x is to say that one is duty-bound to do x. This alters

significantly the way in which ethics is conceived as a practical discipline.

Medicine, in particular, is no longer an apt model. One must look to a

different discipline. Historically, following the tradition of Christian ethics,

this has been jurisprudence. Accordingly, one understands moral stand-

ards, the standards of right and wrong, as analogous to the laws of a

community that regulate its members� conduct. Thus, just as a jurispruden-

tial study of the laws of a community yields conclusions about what actions

its members are legally obligated to perform, so on a deontological concep-

tion of ethics, the study of what it is right and wrong to do in the conduct of

life yields conclusions about what actions a person is duty-bound to per-

form. And just as the determination of what actions a community�s laws

What is ethics? 15

obligate its members to do does not depend entirely on determining what

public or even private good is realized by the observance of those laws, so

too the determination of what actions moral standards bind a person to do

does not depend entirely on determining what good would be realized by

their observance. These standards have authority in practical thought in

virtue of the authority of their source, just as a community�s laws have

authority in virtue of the authority of the legislator or legislative body that

enacted them. And in either case they have such authority independently of

the ends and interests of those whose conduct they regulate.

Plato�s critics believe the Republic�s core problem is inherent in a teleo-

logical conception of ethics. A deontological conception, they think, avoids

the problem. The reason they think so is plain. If the chief conclusions of

ethics are prescriptions about what one ought to do in the sense of being

duty-bound rather than being well-advised in view of certain ends and

interests, then no fact about the advantages or benefits one would gain

from violating a duty of justice or honesty in a given situation challenges

the truth of the prescription that one ought to do the just or honest thing in

that situation. Its truth is unchallenged by such facts since none of them is

relevant to whether one is duty-bound to do the just or honest thing. All

such facts, that is, are consistent with one�s being duty-bound to do it.

Plato�s critics, then, treat the Republic�s core problem as resting on amistake

because they believe the teleological conception of ethics it presupposes is

false. The problem, however, is deeper than they recognize. A deontological

conception of ethics does not avoid it.

6. The problem of deontology

Consider again the problem as it arises from our example of your finding a

lost purse containing a huge wad of cash. When, having found this purse,

you wonder what you ought to do, your question, according to Plato�s

critics, can either be about what duty requires you to do or about what

you would be well-advised to do in view of your ends and interests. Ethics,

they would say, concerns the former and not the latter, and therefore, since

it is the latter and not the former that bids you to forsake basic standards of

honesty as guides to conduct, the problem is due to a simple confusion over

the meaning of the question you are asking. Yet this diagnosis is too quick.

The problem, remember, arises when the question leads you to search for a


good reason to be honest, and if a deontological conception of ethics avoids

this problem, as Plato�s critics believe, then either you must have such a

reason just by virtue of your having a duty not to take what doesn�t belong

to you or your search for such a reason is itself amistake. Either, that is, your

having a duty to do the honest thing is itself a good reason to do it, or you

don�t need to search for such a reason to recognize the authority that basic

standards of honesty have in practical thought. Neither of these alterna-

tives, however, is free of difficulties. Quite the contrary, both are open to

serious objection. Neither, then, allows a deontological conception of ethics

to escape from the Republic�s core problem.

Thus suppose Plato�s critics took the first alternative. Suppose, that is,

they maintained that you have a good reason not to take the cash from the

purse just in virtue of your having a duty not to take it. Your having a duty to

do something, they might say, is itself a good reason to do it. But on what

grounds could they defend this view? �Well,� theymight argue, �as we have

pointed out, if you have a duty to do something, if you are duty-bound to do

it, then you ought to do it, and plainly itmakes no sense to say that someone

ought to do something unless he has a good reason to do it.� But this

response would be a nonstarter. It would amount to begging the question.

No doubt, before Plato�s critics drew their distinction between a use of

�ought� that signifies being duty-bound to do some action and a use that

signifies being well-advised to do an action in view of one�s interests and

ends, wemight have accepted, as a general thesis about the use of �ought� to

prescribe action, that to say that someone ought to do x is to imply that the

person has a good reason to do x. But once they draw their distinction,

acceptance of this general thesis requires separate consideration of the two

cases. Hence, they cannot use the thesis to defend their view without first

showing that it holds for each of the specific uses of �ought� they have

identified, particularly, the use that signifies being duty-bound to do some

action. In other words, before they can use the thesis they must first show

that if one is duty-bound to do some action, one has a good reason to do it.

And this just puts them back to square one.

In response to this criticism, Plato�s critics might try a new tack.

�Admittedly,� theymight say, �we could not use the general thesis to defend

our view if the reasons people had to do things were all of one kind. For in

that case it would bewrong for us to assume that saying that someone ought

to do x implies that he has a reason to do x regardless of which sense of

What is ethics? 17

�ought� one uses to say this. But the reasons people have for doing things are

not all of one kind. Specifically, corresponding to the distinction we draw

between the two uses of �ought,� there is a distinction between moral

reasons and personal reasons. The point is that just as ethics, on our con-

ception of it, is concernedwithwhat people ought to do in the sense of what

they are duty-bound to do rather thanwhat theywould bewell-advised to do

in view of their ends and interests, so too it is concerned with what people

havemoral reasons to do rather thanwhat they have personal reasons to do.

Accordingly, saying that someone ought to do something, if one is using

�ought� in the sense of being duty-bound to do it, is to imply that the person

has a moral reason to do it. Or in other words, your having a duty to do

something gives you a moral reason to do it.�

With this response Plato�s critics would clear themselves of the charge of

begging the question. But in doing so, they would be shifting the grounds on

which they hold that a deontological conception of ethics avoids the Republic�s

core problem. They would be giving up the first of the two alternatives we

identified and taking the second. This should be plain, for they would

be arguing, in effect, that to search for good reasons to do the honest thing,

in circumstances such as yours, is to search for good personal reasons, and

such a search, whatever the outcome, has no bearing on the authority that

basic standards of honesty have in practical thought. To think that it does

would be to make the same mistake as the original one of thinking that

asking whether you ought to do the honest thing in these circumstances

challenges the authority of the standards. That is, just as this question could

not challenge the standards� authority if what you were asking was whether

you would be well-advised to do the honest thing in view of your ends and

interests, so too your search for good reasons to do the honest thing could not

challenge that authority since what you would be after would be good

personal reasons to do the honest thing. To challenge the authority of these

standards the question you were asking would have to be whether you were

duty-bound to do the honest thing and the reasons youwere after would have

to bemoral reasons. But to askwhether youwereduty-bound todo the honest

thing is to ask an idle question, and similarly there is no point in searching for

moral reasons to be honest. You have such reasons because the standards

have authority in practical thought, because they define duties that bind you

to do certain actions, and not the other way round. Or so Plato�s critics, in

shifting to the second alternative, would argue.


On this view, one recognizes the moral reasons one has to do certain

actions by recognizing the authority that moral standards have in practical

thought, and not vice versa. How such authority is to be understood, how-

ever, is not immediately clear. An explanation, though, comes directly from

the way Plato�s critics conceive of ethics as a practical discipline. For their

conception of ethics as a practical discipline is modeled on jurisprudence.

Accordingly, they see the standards ofmorality as analogous to the laws of a

community. Such laws have authority over every member of the commun-

ity and, in virtue of that authority, give each member legal reasons to do

what they require him or her to do. By analogy, then, moral standards have

authority over those whose conduct they regulate and, in virtue of that

authority, give them moral reasons to act as the standards direct. In this

way, Plato�s critics can explainmoral reasons as following from and depend-

ent on the authority of moral standards and not vice versa. Thus, on this

explanation, the basic standards of honesty, in virtue of the authority they

have in practical thought, would give you a moral reason to do the honest

thing in the circumstances you faced independently of your having any

good personal reasons to do it. Yet to advance this explanation, Plato�s

critics would have to assume that you belonged to a community in which

the standards of morality, rather than some code of positive law, say, were

the authoritative standards of conduct and in which each member was

subject to the authority of those standards in virtue of his or her member-

ship in the community. And herein lies the difficulty with their view. After

all, it would not be mere querulousness on your part to ask, �What com-

munity is this? And how did I become a member?�

7. The idea of a moral community

The impulse to think of all human beings as joined together in a moral

community almost certainly lies behind the belief that morality has author-

ity in our lives regardless of our having personal reasons to be moral. It is

one source of the powerful attraction that a deontological conception of

ethics has. There is a global community of all human beings, it is frequently

said, a global village, as it were, and a person qualifies as a member of this

village just in virtue of being human. The community�s laws are the univer-

sal standards ofmorality, and themembers have duties and rights according

to these laws. This thought or something like it, let us then suppose, is what

What is ethics? 19

lies behind the way Plato�s critics explain how the authority of moral stand-

ards precedes and certifies moral reasons. Accordingly, the thought would

supply themwith answers to your questions, for it specifies the community

to which they would think you belonged and sets out the conditions of

your membership. But the thought itself would have to be justified

before Plato�s critics could claim to have shown by this explanation that a

deontological conception of ethics avoids the Republic�s core problem.

Communities of different human beings exist all over the globe, and evi-

dence of their existence consists in their written laws, published rules,

territorial markers, governing institutions, financial arrangements, com-

munal celebrations, ensigns and other symbols of communal unity, and

written and oral histories. Yet there seems to be no such evidence of a global

community to which all human beings belong. How then could Plato�s

critics justify the thought that there was such a community? How could

they show that the thought did not merely reflect their aspiration to a

universal morality?

Lacking empirical evidence of such a community, they must turn to

what they affirm as the universal truths about right and wrong that

correspond to the basic standards of morality. On their conception of

morality, these truths are propositions about what fulfills and what viola-

tes one�s duties, since on this conception matters of right and wrong are

matters of what one ought and ought not to do in the sense of what one is

duty-bound to do. These truths, moreover, are universal inasmuch as the

human practices that create duty would be found in any society. These

include such practices as lending and borrowing, promising and consent-

ing, buying and selling, making friends, entering into marriage, establish-

ing a family, offering and accepting aid, and so forth. Since a society that

lacked such practices is scarcely conceivable, one might then infer from

this observation that, even though no moral community of all human

beings had ever been realized, the basic standards of morality nonetheless

constituted a framework for such a community. One might infer, that is,

that because they corresponded to universal truths about right and wrong

that any reflective person would affirm, they represented valid principles

governing all human social relations both within and across real commun-

ities. Such an inference appears to be the best, if not the only way Plato�s

critics could justify the thought on which their understanding of the

authority of moral standards depends.


Still, it falls short of justifying that thought. Although it may appear that

the universal truths about right and wrong that Plato�s critics affirm corre-

spond to standards of conduct that constitute a framework for a moral

community of all human beings, the appearance ismisleading. These truths

may correspond to such standards, but then again they may not. For what

makes them true (if they are true) is the existence in every human society of

practices that create duties, and consequently we would need some further

reason to think they corresponded to standards of morality that constituted

a framework for a moral community of all human beings. Without such a

reason we cannot assume such correspondence and therefore cannot

assume that they correspond to standards whose authority goes beyond

that of custom. The standards they correspond to may just be the social

norms of a conventional morality. These, too, define duties.

Consider marriage, for example. Marriage may be a practice in every

human society, and if it is, then it is a universal truth that being faithless

to your spouse is wrong inasmuch as the duties that marriage creates

include duties of fidelity to one�s spouse. Yet this truth may correspond

only to the social norms that define such duties, norms that differ among

themselves according as the society to whose conventional morality they

belong practices monogamy or polygamy, enforces patriarchal or egalitar-

ian relations among the sexes, permits or prohibits widows to remarry, and

so forth. Consequently, what Plato�s critics represent as the standard of

morality to which this truth corresponds may come to nothing more than

a generalization of these different norms, in which case no standard corre-

sponding to it would have authority in practical thought that went beyond

the authority of custom. The same points, then, apply to the other universal

truths about right and wrong that Plato�s critics affirm. Hence, these truths

do not provide sufficient grounds for justifying the thought on which the

critics� understanding of the authority of moral standards depends.

The problem at the core of Plato�s Republic has traditionally been a prob-

lem about justice. The philosophical study of morality is a study of stand-

ards of right and wise conduct whose authority in practical thought is

determined by reason rather than custom, and the problem is how to

understand the basic standards of justice as having such authority. How

are these standards to be explained as part of morality? Critics of Plato have

insisted that this problem rests on amistake. In their view, no conception of

morality that left open the question of whether morality included these

What is ethics? 21

standards could ever be right. Thus, on the correct conception, as they see it,

the basic standards of justice are paradigmmoral standards. This idea is the

essence of deontology. But in putting it forth deontologists invite the charge

of infectingmorality generallywith the problemoffinding rational grounds

for the authority of standards of justice. That they invite this charge is

obscured by their various efforts to insulate morality, as they conceive of

it, from embarrassing questions: their distinction between two senses of

�ought�, the corresponding distinction betweenmoral and personal reasons,

their supposition of amoral community of all human beings, their appeal to

universal truths about matters of right and wrong. But, as we�ve now seen,

these efforts serve only to postpone the time at which deontologists must

answer the charge. Full insulation of morality from these embarrassing

questions is not possible. To answer the charge, then, they must show that

the authority moral standards have in practical thought, on their concep-

tion of morality, is determined by reason and not custom. Otherwise their

conception comes down to nothing more than a piece of abstract anthro-

pology. Hence, far from avoiding the Republic�s core problem, it faces that

problem writ large.

8. Ethical theories and moral ideals

To answer the Republic�s core problem requires explaining how justice and

honesty qualify as excellences of character. This requires in turn explaining

how acts of justice and honesty are in themselves reasonable, that is, how an

act�s being the just or honest thing to do gives one, by that fact alone, a good

reason for doing it. Developing these explanations is a task of ethical theory,

and one can find among the many theories that philosophers, since Plato,

have put forward a broad range of different explanations. The explanations

that teleological theories offer connect acting justly with the achievement of

the good that is taken to be the ultimate end of right and wise action, the

summum bonum. What this end is varies from one teleological theory to

another, but on any of them, the explanation must be that acting justly and

honestly are necessary means, or perhaps the best means to achieving it.

Deontological theories, by contrast, must offer explanations of a different

kind. Since on these theories the rightness of acting justly and honestly is not

a matter of whether such actions contribute to the achievement of some

end but rather a matter of their conforming to standards that have authority


in practical thought independently of a person�s ends, the explanations they

offer must bring to light some point to acting justly and honestly. Theymust,

in other words, so enlarge our understanding of those standards and their

place in human life that we see a point to our conforming to them. If, instead,

a deontological theory offered no such explanation, it would leave us in the

dark about why we conform to them. It would ask us to take their authority

on faith and to obey it blindly. It would therefore fail to show that they had

authority in practical thought that was backed by reason rather than custom

and would thus fall short of a central aim of ethical theory.

Implicit in every theory�s answer to the Republic�s core problem is an ideal

of human life. Indeed, no ethical theory could be complete if it did not imply

such an ideal. A complete ethical theory not only formulates and system-

atizes the standards of morality, but also justifies them by laying out the

rational grounds of their authority in practical thought. Such justification,

at a minimum, requires explaining conformity to these standards as mean-

ingful conduct, for you would be at a loss to understand how the authority

of these standards could have rational grounds if you could not find any

meaning to your conforming to them. A complete ethical theory, then, as

part of its justification ofmoral standards, explains how conformity to them

is meaningful, and it does so by showing how such conduct contributes to

your realizing an ideal of human life. Ideals, generally, serve to make

actionsmeaningful in our lives. Many, like those of athletic prowess, artistic

creativity, commercial prosperity, romantic love, family togetherness, tri-

umph over the elements of nature, and so forth, give meaning to common

activities of life by presentingmodels of success in those activities. Having a

model or picture of what success in them consists in enables us to see their

pursuit as something important, worthwhile, or fulfilling. To be sure, none

of these ideals serves to make conformity to moral standards meaningful.

While they present models of success in activities that moral standards

regulate, conformity to moral standards is not what success in those activ-

ities consists in, and therefore to explain such conformity as meaningful an

ethical theorymust incorporate an ideal that applies directly to it. Let us call

such an ideal a moral ideal. It is moral ideals, then, that ethical theories

imply in their answers to the Republic�s core problem.

Needless to say, many people in daily life are seldom if ever troubled

about the meaningfulness of their conforming to moral standards. By and

large, they recognize that general conformity to moral standards by the

What is ethics? 23

members of a community is necessary if people are to live together peace-

fully and that conforming to them also brings such personal advantages as

an untarnished reputation and the goodwill of others. Asked, then, the

point of following such standards, they would likely respond by citing one

or another of these advantages and not some moral ideal. Still, there are

times when one�s circumstances invite acts of dishonesty or injustice that

would neither disrupt social life, tarnish one�s reputation, nor cause one to

lose the goodwill of others. Our example of your finding a lost purse con-

taining a huge wad of cash is a case in point. At these times, one realizes, if

one is sufficiently reflective, that the point of one�s conforming to moral

standards cannot be the necessity of such conformity for social harmony or

for maintaining a sterling reputation and the goodwill of others. At these

times, to find meaning in one�s conforming to them, one must seek a fuller

understanding of their place in one�s life, and the search, if successful, leads

one to affirm some moral ideal.

By the same token, then, an ethical theory, if it succeeds in justifying

moral standards, affirms a moral ideal. Its justification of them consists in

laying out the rational grounds of their authority in practical thought, and it

cannot do this without giving meaning to a person�s following them in

circumstances in which neither social peace nor personal advantage

would be harmed by his ignoring them. It cannot, that is, lay out such

grounds without implying a moral ideal. In this regard, an ethical theory

articulates the thinking of a reflective person who finds himself in such

circumstances and who, having been brought up to act justly and rightly,

now wonders whether there is a point to doing so. His thinking might take

any one of a number of different avenues. Each one would correspond to a

different moral ideal guiding his thought. Accordingly, there are a number

of different ethical theories that articulate these different avenues of

thought and that affirm these different ideals.

It will be the project of the next several chapters to examine these differ-

ent theories. The first ones we will examine are teleological. (See Appendix

for a diagram of these.) Afterwards we will take up those that are deontolog-

ical. Once our survey is complete, we will turn to an important twentieth-

century skeptical attack on these theories and the alternative ethics it offers.

Our examination of the latter will lead to general questions about practical

reason and ethical knowledge, whether either is possible and if it is, how

shall we understand it. The final chapter will deal with these issues.


2 Egoism

1. The wise pursuit of happiness

The question that leads us into the study of different ethical theories

concerns the reasons we have to be honest and just in circumstances that

invite dishonesty or injustice without risk of disrupting social peace, tarn-

ishing one�s reputation, or losing the goodwill of others. One thought a

person who was faced with such circumstances might have is that his

happiness is best served in the long run by adhering to the standards of

honesty and justice. �The cash is very tempting,� he might say to himself as

he looked at the wad of bills in the purse he had just found, �but it would be

stupid to take it. The costs and risks involved make it likely to be more

trouble than it�s worth.� The ideal that a person who thought along these

lines would affirm is that of wisdom in the pursuit of happiness. In ethics,

the theory that affirms this ideal is egoism. The popularity of this theory

among people unfamiliar with moral philosophy suggests that no other

theory has more immediate intuitive appeal. The theory, in addition, has a

secure and important place in the history of ethics. Arguably, it is the theory

Plato worked out in the Republic to answer Thrasymachus� challenge to the

value of justice. In any case, it certainly had other champions in the ancient

world. The most noteworthy of these is the great Hellenistic philosopher

Epicurus (341–271 BC). Its place inmodern philosophy is no less prominent.

In the early modern period its defenders included such major thinkers as

Thomas Hobbes (1588–1679) and Benedict de Spinoza (1632–77), and it

continued to receive strong and important support in the eighteenth and

nineteenth centuries. Only in the twentieth century did its vitality begin to

wane, although even today it still has active and influential defenders.

Egoism is a teleological theory. Its fundamental principle is that the

highest good for each person is his or her own happiness. From this princi-

ple it follows that right action consists in looking out for and furthering

25

one�s happiness, wrong action consists in neglecting it. The theory, in other

words, answers all questions aboutwhat a person ought to do by prescribing

the action by which he can best promote his own happiness. Stated so

baldly, the theory sounds like a recipe for selfishness. But it would be a

mistake to interpret it in this way. Selfishnessmeans that one pursues one�s

own interests without regard to the interests of others. A selfish person

thinks only of himself; nobody else matters to him. Egoism, while its

fundamental principle identifies a person�s own happiness as the highest

good for that person, does not prescribe the pursuit of that good without

regard to the interests of others or concern for their well-being. Indeed,

wherever one�s own happiness depends on other people�s goodwill or their

doing well, it would be foolish not to promote their interests or be con-

cerned about their well-being. And given how pervasive such dependence is

in life, it would be especially foolish not to cultivate friendships, help one�s

neighbors, show kindness to strangers, contribute to the good of one�s

community, and so forth. Egoism, then, far from being a recipe for self-

ishness, encourages behavior and attitudes that are just the opposite. Its

defenders advance it in the belief that basic standards of justice, honesty,

kindness, charity, and the like can be seen to have authority in practical

thought that is grounded in reason when one comes to see that a person�s

happiness depends on his following them. Thus the thought that inspires

the theory is that applying intelligence and forethought to the aim of

achieving happiness shows that the best way to realize this aim is to live

one�s life by such standards.

The thought reflects the ideal of wisdom in pursuit of happiness. This is

the ideal that egoism affirms. A mature understanding of the conditions

necessary for achieving happiness and of the practices, skills, and habits

that make it possible over an entire life shows that one could not have any

realistic hope of achieving it if one lived without regard to the interests of

others or the rules they expect one to follow. Wisdom in the pursuit of

happiness means taking account of one�s long-term as well as one�s short-

term interests and one�s limitations and vulnerabilities as well as one�s

powers and opportunities. Accordingly, a wise person does not squander

his resources on immediate pleasures and transitory benefits but rather

saves a good portion of them as protection against hardship thatmight arise

in the future. Nor does he indulge his appetites and emotions to excess,

risking injury to his health or damage to valuableswhose loss hewould later


regret. Rather he resists their pressure and keeps a level head. Similarly,

then, such a person does not ignore others or trample on their interests for

the sake of paltry rewards and temporary advantages. Instead, he cultivates

friends and tries not to make enemies, for he knows that, however prosper-

ous andwell-liked he is now, theremay come a timewhen he needs the help

of friends and to be free of threats from enemies. And he knows, too, that

the comforts and joys of friendship are among the greatest of human

pleasures. Such wisdom is nothing more than common sense, of course,

but it is no strike against an ethical theory that it draws on common sense.

Egoism is such a theory. Its program is to find in these and similar pieces of

common sense a justification for adherence to basic standards of justice and

honesty, kindness and charity, as well as those of thrift and frugality, self-

control and level-headedness.

Clearly, then, egoism sharply opposes a view like Thrasymachus�. Both,

to be sure, affirm the ideal of wisdom in the pursuit of happiness. They

agree, that is, in their identification of happiness as the highest good.

Where they disagree is on the question of what wisdom requires for suc-

cessful pursuit of this good. Egoism, in answering the question, seeks to

justifymorality as a set of universal standards of right andwrong adherence

to which is the best, if not the only means, to achieving happiness.

Thrasymachus� view, by contrast, represents a kind of skepticism about

morality. It denies that there is a set of universal standards of right and

wrong whose authority in practical thought is backed by reason. At most,

Thrasymachus would say, there is a loose set of maxims about always

keeping your promises, telling the truth, helping those in need, and so

forth that the weak would be wise to follow. But the strong, he would

declare, would not, for they can ignore these maxims with impunity and,

as a result, can gain riches they would otherwise have to forgo. On

Thrasymachus� view, in other words, these maxims do not constitute a set

of universal standards that could be the object of ethical theory, nor do any

others. Hence, his view represents a challenge to the very project of ethical

theory, and this challenge is particularly sharp in the case of egoism. Before

taking it up, however, we need to develop further our exposition of egoism,

to examine the chief arguments for its taking happiness to be the highest

good, and to lay out and assess its program for grounding the basic stand-

ards of justice, honesty, kindness, charity, and the like in its fundamental

principle.

Egoism 27

2. The concept of happiness

One thing essential to understanding egoism is clarification of its central

concept, that of happiness. The clarification is essential because the word

�happiness� is ambiguous. On the one hand, it is sometimes used to denote a

mood, which can be short-lived and even momentary. Thus we talk of

moments of happiness or describe events as bringing a little happiness

into our lives. On the other hand, the word is sometimes used to denote a

condition of a person�s life as a whole or at least over a significant stretch of

it. Thus, when we describe someone as finding happiness or achieving

happiness, we have some such condition in mind. �Happiness,� when used

in the first sense, means something like elation or joy. Its opposite is sad-

ness. �Happiness,� when used in the second sense, means something like a

permanent or long-lasting state of well-being and satisfaction with one�s

situation. Its opposite is misery. This second sense, and not the first, is the

sense we intend when we speak of the pursuit of happiness. The pursuit of

happiness is not the pursuit of elation or joy. It is the pursuit of a durable

state of well-being and satisfaction with one�s life. This pursuit is what

egoism takes to be the context of right and wise action, and accordingly it

is in this second sense of �happiness� that the word is used to formulate the

fundamental principle of egoism. This sense is the theory�s central concept.

The concept combines the idea of living well with the idea of being

satisfied with one�s life. Its two elements, then, are well-being and self-

satisfaction. Many philosophical discussions of happiness focus exclusively

on the first element and ignore the second. Indeed, a common assumption

of these discussions is that the concepts of happiness and well-being are

identical. This assumption, however, is mistaken. Happiness is, in part, a

concept of psychology; well-being is not. To say that someone is happy is to

attribute to him a certain attitude toward himself or his life, whereas no

such attitude is attributed to someone in saying that he is living well.

Accordingly, a full account of the concept of happiness includes an element

that captures this attitude. Contrary, then, to what is commonly assumed,

the concept is not identical to that of well-being. Nonetheless, it does not

follow that a philosophical discussion of happiness necessarily goes wrong

if it focuses onwell-being.What does follow is that inmakingwell-being the

focus of such a discussion one must keep in mind that one is focusing on


only a part of happiness. The main question in such a discussion is what

happiness consists in, and the point is that the answer one comes tomay not

be complete if one comes to it from consideration of what a person�s well-

being consists in. Be this as itmay, consideration ofwhatwell-being consists

in is the principal means to determining what happiness consists in.

What a person�s well-being consists in is among the oldest questions in

ethics. Traditionally, the main answers have been pleasure and excelling at

things worth doing. Hedonism is the theory that supports the first answer.

Perfectionism is the theory that supports the second. According to hedonism,

human well-being consists in pleasure and the absence of pain. The more a

person�s life is filled with pleasure and is free of pain the better that life is.

Perfectionism, by contrast, takes human well-being to consist in activity

that is both worth doing and excellently done. The more a person�s life is

filled with such activity and is free of both trivial actions and failures the

better that life is. In short, hedonism measures a person�s well-being by

the quality of his subjective states, whereas perfectionism measures it by

the worth and character of the activities in which he engages. Hedonism

takes well-being as consisting in pleasant and agreeable experiences

unspoiled by painful and disagreeable ones. Perfectionism takes it as con-

sisting in engaging in worthwhile activities and doing well in them.

These different accounts of well-being then apply directly to the question

of what happiness consists in, though each factors in the second element of

the concept of happiness differently. Thus, because satisfaction with one�s

life is itself a kind of pleasure, dissatisfaction with one�s life a kind of pain,

the hedonist�s account can encompass the second element and therefore

give a complete answer on its own. Hedonism, in other words, can hold that

how satisfied one is with one�s life contributes as much to one�s well-being

as it does to one�s happiness, and therefore the two concepts fall together.

By contrast, because one could be dissatisfied with one�s life even though it

consisted of activities that were worth doing and onewas excelling at them,

the perfectionist�s account must be coupled with the second element to

arrive at a complete answer. Perfectionism, in other words, holds that

satisfaction with one�s life is not a factor of one�s well-being, and therefore

the concepts of well-being and happiness need not extend to all the same

people. The person who is never satisfied with his accomplishments no

matter how extraordinary they are may be doing very well in life yet not

have achieved happiness.

Egoism 29

Egoism can draw on either hedonism or perfectionism for an account of

happiness that guides its determination of right action. The theory�s central

concept, in other words, can be given either a hedonistic or a perfectionistic

interpretation. Nonetheless, there is a theoretical advantage to giving it a

hedonistic interpretation. As we just saw, the hedonist�s account of happi-

ness can unify the concept�s two elements, whereas the perfectionist�s

account cannot, and it stands to reason that a theory will be more coherent

if its central concept is unified. In this case, the disunity in the perfectionist�s

account means that one renders egoism an indeterminate theory when one

takes this account as the one that guides its determination of right action.

The theory is not, by contrast, similarly indeterminate when one takes the

hedonist�s account as the one that guides the theory�s determination of

right action. Hence, giving a hedonistic interpretation to the theory�s cen-

tral concept yields the more coherent and therefore the more powerful

version of egoism.

Consider, for example, the situation of someone who has decided to seek

happiness inmusic. She has a talent for the violin, let us suppose, andwould

regret not pursuing it. At the same time, she may be unsure of how high to

aim in this pursuit. What degree of virtuosity should she strive to achieve?

Of course, she can expect to achieve a greater degree the more talent she

has, but even knowing how talented she is, she may still be uncertain. In

principle, she can resolve her uncertainty if she takes the hedonist�s

account of happiness as her guide. Thus she knows that she would get a

great deal of pleasure from playing the violin well, and the better she played

the more pleasure she would get. One might think, too, that the better she

played the more satisfied she would be with her life, so she should simply

strive to be as good a violinist as she can. But things are not that simple.

Satisfaction with one�s life depends on how closely one�s accomplishments

match one�s aspirations, and so, were she to strive to be as good a violinist as

she could, she might end up, as a result of setting such a high standard, less

satisfied with her life than if she had set her sights somewhat lower. As

William James (1842–1910) sagely put it, a person�s self-esteem is equal to

the ratio of his pretensions to his successes, and if his pretensions are

too high, then his self-esteem will drop notwithstanding his successes.1 To

1 William James, The Principles of Psychology, 2 vols. (1890; reprinted New York: Dover

Publications, 1950), vol. I, p. 310.


resolve her uncertainty, therefore, our violinist must weigh the pleasure of

playing the violin well against the pain of not meeting the standard she sets

and adjust that standard upward or downward until she has set her sights on

a degree of excellence whose pursuit promises to yield more pleasure from

playing the violin well and from a sense of accomplishment in doing so

than the pursuit of some higher degree of excellence or the acceptance of a

lower one.

No similar resolution of her uncertainty is possible, however, if she takes

the perfectionist�s account of happiness as her guide. Because the two

elements of the concept of happiness are irreconcilable on a perfectionistic

interpretation, they represent independent dimensions of happiness.

Accordingly, our violinist, in trying to resolve her uncertainty, must treat

howwell she would play the violin as a separatemeasure of happiness from

how satisfied she would be with her accomplishments. But to weigh how

well she would play against how satisfied she would be with her accom-

plishments is to compare apples and oranges. The two are incommensura-

ble. Consequently, it is not possible for her to determine what standard of

virtuosity she should strive to achieve or whether, having set one, to adjust

it upward or downward. Presumably, up to a point, the higher the standard

she sets the more accomplished a violinist she will become, yet at the same

time after a point, her level of accomplishmentwill fall increasingly short of

the standard she sets and as a result her dissatisfaction will grow. Thus,

relative to the contribution that excellence makes to happiness, on the

perfectionist�s account, she should set for herself the highest standard in

the range of those whose pursuit makes her an increasingly excellent

violinist, and relative to the contribution that satisfaction with her accom-

plishments makes, she should set for herself the lowest standard in the

range of those whose pursuit initiates an increasingly wider gap between

her accomplishments and her aspirations. But the former standard is not

likely to coincide with the latter, and if they diverge, there is no basis on

which she can decide between them or strike a compromise. On the

perfectionist�s account, there is no basis for balancing excellence against

satisfaction in determining what standard to set.

The example brings out the difficulties that the disunity in the

perfectionist�s account of happiness creates. It well illustrates the indeter-

minacy that is produced in egoism when one gives its central concept a

perfectionistic interpretation. That the hedonist�s version of egoism is not

Egoism 31

similarly deficient recommends taking it as the theory�s preferred formula-

tion as long as no argument has settled the issue of whether human well-

being consists of pleasure or excelling at things worth doing. For the rest of

the chapter, therefore, let us put the perfectionist�s version of egoism aside

and assume the hedonist�s version as the standard. Perfectionism will reap-

pear in the next chapter as part of a different theory.

3. The primary argument for egoism

The fundamental principle of egoism identifies happiness as the highest

good for an individual. Accordingly, egoism takes happiness to be the

ultimate end of all right action. It is ultimate in the sense that it is pursued

for its own sake and not for the sake of some other end. In addition, on this

theory, it is uniquely ultimate. That is, in right action, happiness, and

happiness alone, is pursued for its own sake. Any other end one pursues is

pursued for the sake of happiness, either directly or indirectly. Any other

end of right action, in other words, is an intermediate rather than an

ultimate end. As such, it corresponds to an instrumental good. A college

scholarship, for example, is an instrumental good. Winning such a scholar-

ship is helpful to gaining the knowledge and understanding that higher

education imparts. Hence, one seeks it for the sake of gaining such knowl-

edge and understanding. One does not seek it for its own sake, for a college

scholarship can have no value apart from the goods one can use it to attain.

Similar points may then apply to knowledge and understanding. They, too,

would be intermediate ends if they were sought for the sake of the greater

enjoyment of arts and literature, science and technology, that they made

possible. In this case, one would seek them for the sake of the pleasure or

satisfaction their exercise brings. Alternatively, though, one might seek

them for their own sake, in which case they would be ultimate rather

than intermediate ends. Typically, then, on the hedonistic interpretation

of happiness, knowledge and understanding count as instrumental goods.

On the perfectionistic interpretation, they count as constituents of the

highest good.

Few people would argue with the proposition that happiness is an ulti-

mate end of right action. People generally regard their own happiness as

something good in itself and not something they pursue for the sake of

some other end. They also regard it as something that sometimes it is right


to promote even though doing so results in frustrating somebody else�s

pursuits. Egoism, then, in taking happiness as an ultimate end of right

action, squares with ordinary opinion. Where it goes beyond ordinary

opinion is in its taking happiness to be the only ultimate end of right action.

For many people would disagree. They think that what makes an action

right, in some cases, is its being done for the sake of another�s happiness

rather than one�s own. Thus, while they might agree with the proposition

that one�s own happiness is an ultimate end of right action, they would

disagree with the proposition that it was uniquely ultimate. So the main

burden of proof on a defender of egoism is to make a convincing case for

this proposition. The success of any argument for egoism�s fundamental

principle largely rests on its meeting this burden.

Themost important argument wasmade famous by Hobbes. It relies on a

theory of humanmotivation that Hobbes put at the foundation of his moral

and political thought. That theory concerns the springs of intentional

action. To explain it, let us first distinguish, as Hobbes did, between inten-

tional actions, like your drawing the blinds when night falls, and reflexive

behavior, like your blinking when a light is suddenly shown in your eyes.2

Intentional actions spring from motives. They are movements (or refrain-

ings frommovement) that you execute to achieve an end provided by some

motive. Reflexive actions are automatic. They are movements that the

activity of your nervous system produces without the interposition of

some motive. The chief doctrine of Hobbes�s theory is that the motive of

every intentional action is at bottom the same. It is the desire to promote

one�s own interests. This is the doctrine of psychological egoism. The theory

advances this doctrine as a basic truth about human nature and so a funda-

mental law of human psychology. Accordingly, every human motive is

reducible to a self-interested desire. Fear, for instance, is at bottom the

desire to avoid something that threatens to harm one. Anger is at bottom

the desire to retaliate against an aggressor in the interest of deterring future

aggression. Love is at bottom the desire to connect with another whose

companionship brings one pleasure. And so forth. Successful reduction to

self-interested desire of the great variety of motives that we commonly

ascribe to people to explain their actions would thus complete Hobbes�s

theory and confirm its chief doctrine.

2 Hobbes, Leviathan, ch. 6, par. 1.

Egoism 33

The doctrine, if sound, furnishes a defender of egoism with a seemingly

powerful argument for egoism�s fundamental principle. In particular, if

sound, it seems strongly to support the proposition a defender must estab-

lish to meet the main burden of proof on such arguments. Let us call this

argument the primary argument for egoism. Intuitively, it draws on the thought

that itmakes no sense to prescribe for someone actions that best promote of

some end unless he or she could have a motive to pursue that end. What

would be the point? Hence, clearly, if it is a fact of human nature that at

bottom the only motive people could ever have is the desire to promote

their own interests, then the only end whose promotion it makes sense to

prescribe as an ultimate end is a person�s own interests, which is to say, his

or her own happiness. Therefore, the only end whose promotion it makes

sense for a fundamental principle of right action to prescribe is the actor�s

own happiness. And this, in effect, is the proposition the defender of egoism

must establish.

The primary argument can be given a more explicit statement, one that

makes clearer how it proceeds from its psychological premisses to its

ethical conclusions. On this more explicit statement, the argument�s basic

premiss is the doctrine of psychological egoism (that the motive of every

intentional action is at bottom the desire to promote one�s own interests). A

second premiss is that if the basic motive of every intentional action is the

same, then the end provided by thatmotive is the only ultimate end of every

intentional action. From these two premisses, it follows that one�s own

interests, which is to say, one�s own happiness, is the only ultimate end of

every intentional action. And since all right actions are intentional actions –

a third premiss – it further follows that the actor�s own happiness is the only

ultimate end of right action. Hence, the highest good for an individual is his

or her own happiness. Hence, from the doctrine of psychological egoism

one can, in a few, simple steps, arrive at the fundamental principle of

egoism.

4. Psychological egoism

Plainly, the weight of this argument lies with its first premiss, the doctrine

of psychological egoism. So the question to ask is: How plausible is it?

Some thinkers have thought it was very plausible if not indisputable.

Others have dismissed it as utterly preposterous. Given such a great


disparity in opinions, one might suspect that the doctrine was understood

differently by the different parties to the dispute. So we would do well to

examine this possibility, considering first how it might be understood as

plausible and whether, on such an understanding, the primary argument

for egoism would still be cogent. Certainly, the doctrine would be plausible

if all it meant were that a person�s intentional actions always sprang from

motives supplied by his own interests. After all, it would be hard to see how

a person could be moved to pursue something if he had absolutely no

interest in it and did not believe that failure to pursue it would lead to the

loss of something he did have an interest in. But if the doctrine came to no

more than that a person�s own interests were the sole source of the motives

of his or her intentional actions, then the primary argument would not be

cogent. Specifically, its basic premiss would be too weak to support the

proposition a defender of egoism must establish for the argument to meet

the main burden of proof on arguments for egoism. The reason is that a

person typically has interests in many things and people besides himself:

his family, his friends, his neighborhood, his country, the associations and

communities to which he belongs, his relationship to God, and so forth.

These interests focus his thoughts and feelings on people and things other

than himself, and accordingly the motives that they supply are motives to

act for the sake of those people and things and not for his own sake.

Take, for example, the interest a person typically has in his family.

Typically, a person regards his family�s welfare as something that is impor-

tant to him absolutely and not just in relation to his own happiness. So his

interest in his family supplies a motive to promote their welfare for their

sake alone. It follows, therefore, that their welfare, and not his own happi-

ness, is the end this motive provides. Consequently, the ultimate end of the

intentional actions that spring from this motive need not be his own

happiness. It may be his family�s welfare only and, indeed, will be if the

actions spring from no other motive. And similar points follow about the

motives that many of the other interests a person has in people and things

besides himself supply. Plainly, then, if the doctrine of psychological ego-

ism came to no more than that a person�s interests were the sole source of

the motives of all of his or her intentional actions, one could not infer from

it that the ultimate end of every intentional action was the actor�s own

happiness. Yet without this inference the primary argument for egoism

breaks down. Without it one cannot go on to infer that the ultimate end

Egoism 35

of every right action is the actor�s ownhappiness, and this is the proposition

a defender of egoism must establish to meet the main burden of proof on

arguments for egoism.

It would thus appear that the doctrine, to be strong enough to support

this proposition, must be understood to mean something more than that a

person�s intentional actions always spring from motives supplied by his

own interests. It must also be understood to imply that none of the motives

a person�s interests supply him with has force independently of the motive

to promote his own happiness. But now the question is whether the doc-

trine, on this understanding of it, is at all plausible. Is it at all plausible to

think that none of our interests in people and things besides ourselves

supplies us with motives whose force is independent of the motive to

promote our own happiness? Is it plausible, for instance, to think that our

interests in our friends and our family do not supply us with such motives?

Admittedly, if these interests were like the interest an individual stock-

holder typically has in the companies in which she invests, then it would

be plausible to think that they never supplied us with such motives. For in

this case, they would derive entirely from the interest we have in our own

happiness, just as the stockholder�s interest derives entirely from her inter-

est in herfinancial well-being. But formost of us the interests we have in our

friends and our family are not like the interest of the typical individual

stockholder in the companies in which she invests. They are not just

interests in the utility of our friends� and our family�s prosperity for our

own happiness. Rather they are also interests in our friends� and family�s

welfare as something important to us in itself. For this reason it is implau-

sible to deny that they supply uswithmotiveswhose force is independent of

themotive to promote our own happiness. The doctrine, therefore, must be

judged as similarly implausible when it is understood to exclude the possi-

bility of such motives.

The upshot of this examination then is that the primary argument for

egoism runs into serious difficulty once we consider how the doctrine of

psychological egoism is best understood. In a word, the argument is caught

in a bind. If the doctrine is understood in a way that makes it strong enough

to support the proposition that the ultimate end of every intentional action

is the actor�s own happiness, then it is rendered implausible, and conse-

quently the argument fizzles out from a faulty premiss. If instead the

doctrine is understood in a way that makes it plausible, then it is rendered


too weak to support this proposition, and consequently the argument

breaks down at its first inference. The question, then, is whether there is

any way out of this bind. On balance, the most hopeful strategy is first to

construe the doctrine as strong enough to support the proposition and then

to argue for its plausibility. But successful pursuit of this strategy has yet to

occur. Invariably, attempts to show that the doctrine is plausible, given the

way it�s being construed, fall short of explaining away (or explaining as

wholly self-interested) the motives of friendship, paternal and filial love,

civic pride, community spirit, religious devotion, human kindness and

sympathy, and the like that sorely test its credibility.

5. An alternative argument for egoism

A variation on this strategy is to construct a new argument for egoism that

proceeds along similar lines as the primary argument but with a different

psychological doctrine as its basic premiss. The theory of humanmotivation

that best fits this strategy is psychological hedonism. Its chief doctrine is

that the motive of every intentional action is at bottom either a desire for

pleasure or an aversion to pain. Psychological hedonism is one of the oldest

and most influential theories of human motivation in the history of philos-

ophy and psychology. The famous opening lines of Jeremy Bentham�s great

work, An Introduction to the Principles of Morals and Legislation, capture it

succinctly: �Nature has placed mankind under the governance of two

sovereign masters, pain and pleasure. It is for them alone to point out

what we ought to do as well as to determine what we shall do.�3 Bentham

(1748–1832) himself, we should note, did not use the theory to ground the

fundamental principle of egoism. He thought it supported a different prin-

ciple, and later wewill find it instructive to see why. Butmany philosophers

have thought that the theory does support egoism, and the argument they

favor is worth examining to determine whether it makes a better case for

egoism than the primary argument.

While the argument�s basic premiss, the doctrine of psychological

hedonism, is different from that of the primary argument, its second prem-

iss is the same. In addition, the argument follows, at the beginning, the

same pattern of inferences. So its first few steps parallel those of the earlier

3 Bentham, An Introduction to the Principles of Morals and Legislation, ch. 1, par. 1.

Egoism 37

argument. Thus itsfirst two premisses combine to yield the proposition that

the ultimate end of every intentional action is to experience pleasure or

escape from pain. And from this proposition and the same third premiss as

the primary argument�s, which is that all right actions are intentional

actions, it follows that the ultimate end of every right action is to experience

pleasure or escape from pain. At this point, however, the two arguments

diverge.Where the primary argumentmoves quickly to its final conclusion,

the argument from psychological hedonism takes more steps. The reason

for this difference is worth considering.

The primary argument moves quickly at this point to its conclusion

thanks to an earlier simplification. The simplification is to equate the pur-

suit of one�s interests with the pursuit of happiness. Yet, as our criticism of

it makes clear, this equationmasks a serious problem. Specifically, it masks

an ambiguity in the doctrine of psychological egoism that, once detected,

exposes the argument as having either a faulty premiss or an unsound

inference. In the argument from psychological hedonism, by contrast, no

corresponding simplification occurs. For the corresponding simplification

would be to equate the pursuit of pleasure (and freedom frompain) with the

pursuit of happiness, and this equation is clearly false. Imprudent and self-

indulgent pursuits of pleasure are as sure a road to unhappiness as any. In

other words, while the ultimate end of every right action is either pleasure

or freedom from pain, it is not pleasure nomatter howmuch pain one ends

up suffering as a result or freedom from pain nomatter howmuch pleasure

one ends up having to forgo. The person who immerses himself in imme-

diate pleasures heedless of future, deleterious consequences to his health

and well-being is not acting rightly. The argument, then, includes an addi-

tional premiss whose point is to deny that such imprudent and self-

indulgent actions are right actions.

This fourth premiss is that an act is right if the actor could not obtain, by

some other action,more ofwhat he is pursuing or less of what he is trying to

escape from and if nothing else besides what he is pursuing or the absence

of what he is trying to escape from is an ultimate end of his intentional

actions. The premiss gives a formal condition on the rightness of an action

within a teleological theory. As such, it represents a requirement of reason,

specifically, in this case, a requirement of economic reason. If one is pursu-

ing, as an ultimate end, something of which one can enjoymore or less, it is

not rational to opt for less when one knows one can get more without


additional cost. When this requirement is combined with the identification

of one�s experiencing pleasure or escaping from pain as the ultimate end of

every right action, it follows that the only ultimate end of every right action

is one�s experiencing asmuch pleasure and as little pain as one can. In other

words, given the hedonistic account of happiness, one�s own happiness is

the only ultimate end of every right action. It is, therefore, the highest good

for an individual. Thus, using a somewhatmore complicated argument than

the primary argument, one can base the fundamental principle of egoism

on the doctrine of psychological hedonism.

6. Psychological hedonism

Needless to say, this argument would be no more successful than the

primary argument if the doctrine of psychological hedonism were as prob-

lematic as the doctrine of psychological egoism. So the main question is

how well it holds up on examination. How plausible is it to think that the

motive of every intentional action is at bottom either a desire for pleasure or

an aversion to pain? This question might not seem to pose much of a

challenge. Indeed, it is tempting to regard the doctrine as a plain truth.

After all, since one can characterize anymotive as either a kind of desire or a

kind of aversion, the question amounts to asking about the plausibility of

the thesis that every desire is a desire for pleasure and every aversion an

aversion to pain. And this thesis may seem beyond dispute. No one, you

might think, could have a desire thatwas not in essence a desire for pleasure

or an aversion that was not in essence an aversion to pain, at least if

�pleasure� is taken as the general name of every sort of agreeable experience

and �pain� the general name of every sort of disagreeable one. As John Stuart

Mill (1806–73) once wrote, �to desire anything except in proportion as the

idea of it is pleasant is a physical and metaphysical impossibility.�4

Yet though the doctrine may seem to be a plain truth, it is hardly such.

What makes it seem plainly true is a common confusion. When you have a

desire for something, you typically anticipate getting pleasure from it. Let

us even suppose that you invariably anticipate such pleasure when you

desire something. This supposition is certainly plausible, especially given

the plausibility of supposing that when you desire something, you

4 Mill, Utilitarianism, ch. 4, par. 10.

Egoism 39

anticipate having the desire satisfied and that such satisfaction is itself a

source of pleasure. Nonetheless, it does not follow that the pleasure you

anticipate getting is always the true object of your desire. It maymerely be a

welcome extra. To be sure, sometimes a person may desire something, an

apple, say, just for the pleasure of its taste, in which case the pleasure he

anticipates is also the true object of his desire. But sometimes a personmay

desire an apple for its nutritional value, in which case, though he antici-

pates the pleasure of its taste, this pleasure is not what he is after. So it is not

the object of his desire. More important, still, is the observation that the

pleasure you anticipate getting when you have a desire cannot be the object

of that desire in the case in which the source of the pleasure is the satisfac-

tion of the desire. The reason for this is internal to logic of desire. Your

desire for something is satisfied when you get the thing you desire.

Therefore, the object of your desire – the thing you desire –must be distinct

from the satisfaction of the desire, since the satisfaction presupposes the

object. But this means that the pleasure you get from the satisfaction of the

desire cannot be the desire�s object, since this pleasure, too, presupposes

the object. The common confusion thatmakes the doctrine of psychological

hedonism seem plainly true is due to a tacit and faulty assumption that the

two are identical.

The observation that the object of a desire cannot be the pleasure one

gets from the satisfaction of that desire is the key to a powerful criticism of

psychological hedonism that Joseph Butler (1692–1752) first made in his

sermons on human nature.5 Butler used this observation to support a dis-

tinction he drew between the general desire for one�s own happiness, what

he called self-love, and particular desires for food, clothing, riches, the

welfare of one�s children, the good opinion of one�s fellows, and so forth,

which he identified as particular appetites and passions. Because the sat-

isfaction of one�s particular desires is essential to happiness, Butler argued,

self-love involves self-reflection. A person who seeks happiness necessarily

reflects on how best to satisfy her particular desires. Thus, the pleasures of

such satisfaction are the concerns of self-love; the most felicitous combina-

tion of them over a whole life is its object. This point nicely fits with the

important observation that the object of a desire cannot be the pleasure one

gets from the satisfaction of that desire. Such pleasure, because it is

5 Butler, Fifteen Sermons Preached at the Rolls Chapel, sermon xi.


essential to happiness, is obviously the object of a human desire, and

Butler�s point is that self-love is that desire, while the particular desires,

whose satisfaction is the object of self-love, have different objects. In some

cases, no doubt, these are experiences of sensory pleasure, like the taste of

chocolate or the smell of roses. But in others they are not. Rather they are

things such as wholesome food, well-made clothing, gems and other riches,

the welfare of one�s children, a good reputation, and the like. Defenders of

psychological hedonism have to deny, of course, that these things are ever

the true objects of particular desires. They have to maintain instead that all

particular desires are desires for experiences of sensory pleasure (or the

absence of sensory pain). But it is hard to see any grounds for so restricting

the objects of particular desires. It is hard, that is, to see any grounds on

which the doctrine of psychological hedonism can be sustained.

We can summarize Butler�s criticism of the doctrine briefly. Happiness

consists, in large part, of the pleasures that arise from the satisfaction of

one�s particular desires, and those desires are satisfied when one gets the

things one desires. The desire for happiness is self-love. Its object is the

pleasure that comes from satisfying one�s particular desires. The objects of

those desires are various, but in no case can the object of a particular desire

be the pleasure one gets from its satisfaction, for such satisfaction presup-

poses a distinct object. If psychological hedonismwere true, then the object

of every particular desire would have to be the experience of a sensory

pleasure (or the absence of a sensory pain), for it is the only thing consistent

with both the doctrine of psychological hedonism and Butler�s observation

that the object of a particular desire cannot be the pleasure one gets from

the desire�s satisfaction (or the absence of the pain one experiences from

the desire�s frustration). But to restrict the objects of particular desires

in this way appears arbitrary. To hold, for instance, that a person�s desire

for the welfare of his children must at bottom be, say, the desire for

experiences of sensory pleasure that result from seeing his children flourish

seems very far-fetched.

Butler�s criticism, in bringing out the implausibility of psychological

hedonism once the desire for happiness is distinguished from other desires,

parallels our earlier criticism of the doctrine of psychological egoism. That

criticism similarly showed, by distinguishing between a person�s interest in

his own happiness and his interests in people and things outside of himself,

that the doctrine of psychological egoism is implausible when it is

Egoism 41

understood to be strong enough to support the proposition that the ulti-

mate end of every intentional action is one�s own happiness. On this under-

standing, the doctrine, as we saw, implies that none of the many external

interests a person has, such as his interests in his family and friends,

supplies him with a motive whose force is independent of the motive to

promote his own happiness, and to deny that any of these interests supply

such a motive is like denying that any of a person�s particular desires has as

its object something other than the experience of pleasure or the absence of

pain. The denials in both cases are implausible.

This parallel between the two criticisms suggests, moreover, that the

relation between a person�s particular desires and his general desire for his

own happiness, his self-love, from which Butler drew his criticism is a good

model for understanding the relation between a person�s external interests

and his interest in his own happiness, his self-interest. Accordingly, the

satisfaction of a person�s external interests contributes importantly to his

happiness, and his self-interest is thus partly an interest in the satisfaction

of these external interests. Themotives they supply can therefore be under-

stood as distinct from the motive of self-interest. Attempts by defenders of

psychological egoism to show the contrary, that the motivational force of

the former must derive in every case from that of the latter, invariably fail.

On a more plausible theory of human motivation, a person develops inter-

ests in people and things besides himself as he becomes, in growing up,

increasingly engaged with people and things around him. At some point,

perhaps very early in his development, he comes to reflect on these inter-

ests and the importance to him of their satisfaction and, in consequence,

develops an interest in his own happiness. At the same time, that the

development of his external interests precedes the development of his

self-interest means that the motives they supply have force independently

of the motive it supplies, and there is no reason to think that they lose their

independent force with the development of self-interest and the incorpo-

ration into it of the external interests that supply these motives. The

implausibility of psychological egoism, like the implausibility of psycho-

logical hedonism, can be traced, then, to a dubious view of the relation

between the desire for happiness and the desires and interests whose

satisfaction is essential to happiness.

The lesson in these criticisms is that no argument for egoism is likely to

succeed if its premisses deny the fundamental diversity of human motives.


Themost plausible theories of human psychology affirm such diversity and,

in consequence, affirm the possibility of more than one ultimate end of

intentional action. Since each of these ends can be regarded as a candidate

for the highest good, a direct argument for egoismmust present reasons for

regarding one�s happiness as more desirable or worth pursuing than any of

these other ends. It cannot, that is, consistently with the assumption of the

fundamental diversity of human motives, preclude these other ends from

even being considered on the grounds that at bottom the ultimate end of

every intentional action is the same. It must instead make the case for

egoism on the merits of having one�s own happiness and not one of these

alternatives be the ultimate end of one�s actions. It must, in other words,

make the case for its always being right to act on self-love and never right to

suppress self-love in the interest of acting on some other motive that con-

flicts with it. The prevalence of self-love in human life and the difficulty of

seeing a person�s preferring the satisfaction of his own interests to the

satisfaction of others� as unreasonable make clear why a person�s own

happiness is a leading candidate for the highest good and may also suggest

how such an argument might go. At the same, it is hardly obvious that such

an argument would succeed in the end. The burden of proving that one�s

own happiness is uniquely the ultimate end of right action remains a large

obstacle. The important thing, though, is that, being consistent with the

fundamental diversity of human motives, it would in this respect be an

advance over the primary argument for egoism and the argument from

psychological hedonism.

7. The Hobbesian program

Both the primary argument for egoism and the argument from psycholog-

ical hedonism take the form of a proof of egoism�s fundamental principle.

Their failure can of course be an inspiration to search for a better argument

of this form. Alternatively, though, one could drop the idea of finding such

an argument and look instead for a less direct defense of egoism.

Specifically, one could look for such a defense to the theory�s program of

grounding the basic standards of justice, honesty, kindness, charity, and the

like in its fundamental principle. The thought here is that the success of this

program would itself be an argument for accepting the theory, notwith-

standing the difficulties of finding a proof of its fundamental principle.

Egoism 43

Indeed, if it should turn out that there is no ethical theory whose funda-

mental principle or principles are susceptible of such proof, then the suc-

cess of this program could be a powerful and perhaps decisive argument for

the theory. Many systems of thought are developed from postulates, which

are the unproved �first truths� of a system and from which the system�s

major results are derived as theorems, and an important measure of the

soundness of such systems is what propositions are derivable within it as

theorems. To be sure, egoism lacks the rigor one expects of such systems.

But this should not preclude it from being judged by the same measure of

soundness. Consequently, if the theory�s program of grounding basic stand-

ards of right and wrong in its fundamental principle were successful, one

could, by appealing to this measure, make a persuasive case for it.

The program received its definitive statement in Hobbes�s Leviathan. In

two crucial chapters of that work, Hobbes produced derivations of more

than a dozen rules of right action, which he called laws of nature, and these

derivations are a useful guide to carrying out what I shall call the Hobbesian

program, the program of grounding what are commonly taken as the basic

standards of right and wrong in egoism�s fundamental principle.6 Equally

useful are the important reflections on the natural conditions of human life

that Hobbes offered in the chapter immediately preceding his derivations.7

Hobbes, in this earlier chapter, drew a very bleak picture of what human life

would be like if men and women did not follow the laws of nature in their

conduct toward each other. �Solitary, poor, nasty, brutish, and short� were

the words he used to sum up his description.8 The description, by itself,

seems excessively harsh, but Hobbes supported it by showing how such

harsh conditions would eventuate given a few natural facts – or what he

took to be facts – about human beings and their habitat and given his

hypothesis of conduct ungoverned by the principles of right action. His

point was that, since the facts are unchangeable, human beings must

govern their conduct by these principles to avoid the short and miserable

life that would otherwise eventuate. He further thought that they needed

rulers and a state to enforce the principles. But it is not necessary to go

into his political theory to understand his derivations and so to use them as

6 Hobbes, Leviathan, chs. 14 and 15. 7 Ibid., ch. 13.8 Ibid., ch. 13, par. 9. Spelling in this and subsequent quotations from Hobbes�s Leviathan

changed to conform to modern English usage.


guides to carrying out the Hobbesian program. Grasping his point about

everyone�s needing to follow the standards he derived to avoid a short and

miserable life should be sufficient. His derivations can then be understood as

conditioned on the natural facts he adduced to support this point.

Three are crucial. First, there is the predominance, among themotives to

which human beings are subject, of self-love and the desire for the neces-

sities of life. For Hobbes, this fact was the germ of the theory of human

motivation he put at the foundation of his moral and political thought, and

thus one might, in view of our criticism of that theory, wonder whether he

had really gotten hold of a fact in this case. But regardless of how Hobbes

conceived of the predominance of self-love among humanmotives, it can be

understood apart from the theory and thus seen to be both free of the

theory�s special problems and unlikely to excite controversy. For it should

be uncontroversial – and this is the first fact – that in human beings self-love

and the desire for the necessities of life tend to dominate the altruistic

motives to which human beings are also subject, and their domination is

greater the more socially distant are the potential beneficiaries of the altru-

ism that the latter motives prompt. The second fact is that of scarcity in the

necessities of life. The earth, Hobbes held, is not so hospitable a habitat for

human beings that they can get the things they need to nourish themselves,

protect themselves from the elements, maintain their health, care for their

small children, and so forthwithout expending any effort orworrying about

coming up short. Finally, the third fact is that of a certain equality among

human beings. People are all equal, Hobbes said, in that they are all vulner-

able to lethal attack and mortal injury. This is obviously a weak criterion of

human equality, but Hobbes�s derivations required nothing stronger. For

convenience� sake, let us refer to these three facts as the facts of equality,

scarcity, and predominating self-love.

Hobbes conditioned his derivations on these facts by showing that they

define circumstances in which all human beings would be better off collec-

tively if, in their conduct towards each other, they followed standards of the

kind he derived than if they followed no such standards. His argument

begins with the fact of scarcity. Thus, because of scarcity, people would be

better off collectively if they got along with each other andworked together

than if they isolated themselves from each other and tried to scrape by on

their own. Such cooperation would be more beneficial for them both

because it would reduce the friction and potential for conflict that

Egoism 45

competition for scarce goods inevitably creates and because it would,

through collaborative enterprise, increase the overall supply of goods that

human beings need to live a tolerable life. Unfortunately, because of pre-

dominating self-love, one cannot expect people to get alongwith each other

or work together solely out of the goodness of their hearts. If they were

naturally more motivated by kindness and generosity than by self-love and

the desire for the necessities of life, then of course they could get along and

work together without giving any thought to whether such conduct was in

their self-interest. But since the opposite is true, they cannot. Hence, to gain

the fruits of cooperation they must recognize and follow standards of con-

duct general adherence to which would enable them to get along and work

together in ways that were mutually beneficial. Finally, then, because of

equality, every human being is similarly situated with respect to these

circumstances. Thus, no human being would be better off if people didn�t

get along or work together. The reason why should be clear. The greater

hostility among people – the increased conflict that competition for scarce

goods inevitably creates – significantly increases everyone�s risk of being

the victim of a lethal attack. Hence, everyone would be better off if they

recognized and followed standards of conduct general adherence to which

enabled them to get along and work together than if they ignored them.

While these considerations show that to live in circumstances in which

people recognize and follow such standards is to everyone�s advantage, it

still remains an open question whether someone who lives in such circum-

stances always best promotes his happiness by adhering to these standards.

The question remains open because it is possible for someone who lived in

such circumstances to enjoy the benefits of people�s recognizing and fol-

lowing standards general adherence towhich enables them to get along and

work together without following the standards himself. It is possible, that

is, for someone to get a free ride, for a person could evade the standards

without thereby affecting whether others continued to adhere to them, and

if they did, then the benefits of their adherence would still be available to

all, including him. This possibility plainly threatens the Hobbesian pro-

gram, for if someone living in such circumstances could best promote his

happiness by selectively rather than strictly adhering to these standards,

then for that person, according to egoism, adhering to them is sometimes

the wrong thing to do. The standards, in other words, would lack authority

in practical thought that is backed by reason, since at these times following


them is inconsistent with what reason requires, namely, promoting the

highest good. The theory, therefore, would have fallen short of vindicating

them as the standards of morality. Hence, to ground them in the theory�s

fundamental principle, one must further show that no one who lives in

circumstances in which people, for the most part, strictly adhere to these

standards could best promote his happiness by selectively rather strictly

adhering to them. Finding an argument that shows this is a make-or-break

challenge, so to speak, for the Hobbesian program. It is the challenge posed

by the free-rider.

The challenge did not escape Hobbes. He dealt with it as part of his

derivation of the natural law of justice, which he construed as prescribing

that one keep one�s word. For reasons having to do with his political theory,

he took keeping one�s word as the primary act of cooperation by which

people reduce the conflict among themselves that competition for scarce

goods inevitably creates. So in answering the challenge, he considered

whether one could, in circumstances in which others strictly kept their

word in their dealingswith one, best promote one�s happiness by selectively

rather than strictly keeping one�s word in one�s dealings with them. His

answer, though, can be easily adapted to answering the broader challenge

to the entire Hobbesian program and not just to his derivation of the law of

justice. Thus, on the Hobbesian program, every act of following standards

general adherence to which enables people to get along and work together

counts as an act of cooperation, no matter which particular standard is

being followed. Accordingly, let us define cooperation as following such

standards of conduct. The question to be considered, then, is whether one

can, in circumstances in which others always cooperate with one, best

promote one�s happiness by selectively rather than strictly cooperating

with them.

Hobbes�s answer appeals to the risks and costs of being excluded from

the cooperative arrangements that general adherence to the standards

makes possible. The costs of such exclusion, Hobbes maintained, are enor-

mous. They include not only one�s being excluded from sharing in the

surplus of goods that result from people�s working together in collective

enterprises but also one�s becoming a target of their hostility and fair game

for predatory invasions. In short, to be excluded from these arrangements is

to be made a pariah, and the life of a pariah is wretched indeed. No benefits

could be worth such a cost. It is of paramount importance, then, that one

Egoism 47

avoid incurring it. To do so, one must maintain the trust of others. Trust is

what cements the relations among people on which cooperative arrange-

ments for their mutual benefit depend. If one loses the trust of others, one�s

relations with them become ruptured. Cooperation with them breaks

down. And since the way people come to trust each other is by their

manifesting certain dispositions of character in their dealings with one

another, a person could not reasonably expect to maintain the trust of

others if he did not cultivate in himself these dispositions. Call these the

dispositions of a cooperator. They include dispositions like justice, honesty,

kindness, modesty, fair-mindedness, the willingness to compromise, inter

alia, and to cultivate them in oneself requires that one cooperate strictly

with others and forbear from cheating and bullying them when the oppor-

tunity for profiting by such actions arises. Selective cooperation, in other

words, is inconsistent with cultivating in oneself the dispositions of char-

acterwhosemanifestation in one�s dealingswith others elicits and strength-

ens their trust. Hence, selectively cooperating with others without losing

their trust would be possible only if one were able to fool them about one�s

character. And one cannot, Hobbes insisted, count on being able to do this.

Hobbes, it is important to note, did not deny that one could sometimes

profit more by cheating others than by cooperating with them. He did not

deny that a person might gain greater happiness by selectively rather than

strictly cooperating with others. His point, rather, was that if one did gain

greater happiness through selective cooperation, one would have gained it

by luck and not solely by design. For success at cheating others, he argued, is

contingent on their never coming to know that one is untrustworthy, which

is to say, that one lacks the dispositions of a cooperator, and one cannot

bank on their remaining ignorant of these deficits in one�s character. Such

ignorance depends on too many factors outside one�s control. So selective

cooperation, the willingness to cheat others when the opportunity for

profiting by doing so arises, is a risky way of dealing with them. And given

what one risks, the loss of others� trust, the consequent exclusion from

cooperative arrangements, and the pariah status such exclusion entails, the

greater happiness one might gain from cheating can hardly be worth this

risk. Thus, strict cooperation with others, Hobbes concluded, is the wisest

course for one to take in circumstances in which others are, for the most

part, following the same course. This result, then, completes the defense of

the Hobbesian program.


8. Troubles with the Hobbesian program�s derivations

Hobbes�s answer makes a strong case for cultivating in oneself the disposi-

tions of a cooperator. At the same, one may still wonder how successful his

answer is in showing that cheating someone is never the right thing to do in

circumstances inwhich people, for themost part, are cooperatingwith each

other. Clearly, if Hobbes�s answer is that cheating is never the right thing to

do in such circumstances because it always involves a risk of being found

out that is too great to be worth whatever benefits one can get from cheat-

ing, then the answer is unsound. Its premiss about the riskiness of cheating

is unrealistic. Granted, one seldom finds oneself in circumstances in which

one can cheat and the risk of being discovered is negligible. But to assert

that people never find themselves in such circumstances is a gross over-

statement. Sometimes one finds a lost purse containing a huge wad of cash

when nobody else is around. It is possible, though, to take Hobbes�s answer

to be based on a different premiss. One can take it to be based on the premiss

that no one who cultivates in himself the dispositions of a cooperator can

ever regard cheating, no matter how small the risk of being found out, as a

more profitable course of action than cooperating. Hobbes�s answer, then, if

it is understood as based on this premiss, is that cheating is never the right

thing to do because it is always inconsistent with one�s having the disposi-

tions of a cooperator. Either one has these dispositions, in which case

cheating is not an option, or one does not have them, in which case the

risk of being made a pariah if one fails to cultivate them in oneself is too

great to be worth whatever benefits come from cheating. And this more

sophisticated version of Hobbes�s answer is not so clearly unsound.

Nevertheless, it is still open to criticism. A disposition of character is only

a strong tendency to choose to act in certain ways. It is not a force that

invariably produces such action in circumstances in which that sort of

action is expected of someone who has the disposition. Consequently, a

person�s cheating on rare occasion would not necessarily imply that he

lacked any of the dispositions of a cooperator. Suppose, then, a person

who had cultivated in himself these dispositions found himself, unexpect-

edly, in circumstances in which cheating would add considerably to his

wealth and the risk of being discovered was negligible. If the person could

cheat in these circumstances and still have the dispositions of a cooperator,

Egoism 49

then, plainly, cheating would best promote his happiness. In retaining

these dispositions, he would be at no greater risk of failing to manifest

them in subsequent dealingswith others than if he cooperated. Hewould be

at no greater risk of being made a pariah. Cheating, therefore, would be the

right thing for him to do in these circumstances, on the definition of right

action that egoism endorses. Hence, this more sophisticated version of

Hobbes�s answer would also be unsound. It, too, would fail to show that

cheating is never the right action in circumstances in which people, for the

most part, are cooperating with each other. It, too, would be based on a

faulty premiss, in this case the premiss that cheating is always inconsistent

with having the dispositions of a cooperator.

A defender of this version of Hobbes�s answer would no doubt object.

Honesty, hewould point out, is a disposition of character a person could not

possess if he chose to cheat in these circumstances, and honesty is a para-

digm of the kind of disposition whose manifestation in a person�s dealings

with others elicits and strengthens their trust. But this objection simply

assumes what a defender of Hobbes�s answer must show, for once the

question of whether cheating is always inconsistent with possessing the

dispositions of a cooperator surfaces, we can no longer assume that the only

sort of dispositions that qualify as dispositions of a cooperator are virtues

like honesty and justice. To be sure, on Hobbes�s answer, one does well in

cultivating these virtues in oneself. Manifesting them in one�s dealings with

others is certain to gain and maintain their trust, and one�s happiness

depends on this. But one would do even better if one cultivated in oneself

dispositions that, on the one hand, were not so rigid as to keep one from

taking advantage of circumstances in which cheating would bring large

benefits and the risk of being discovered was negligible but, on the other,

were sufficiently sensitive to the need tomaintain others� trust as to qualify

as the dispositions of a cooperator. Given that a personmight cultivate such

dispositions in himself, cheating need not always be inconsistent with

being a cooperator.

But is it possible for someone to cultivate such dispositions in himself?

Could someone be this flexible about following standards of conduct adher-

ence to which he regards as important to achieving happiness? Consider,

as an analogy, the dispositions of self-discipline. A person who is self-

disciplined is able to keep to a regimen: a diet, a budget, a schedule, and

so forth, yet at the same time she need not be so rigid about keeping to a


diet, say, that if a once-in-a-lifetime opportunity to have a fabulous dinner

prepared by a world-renowned chef came along, she would pass it up to

keep to her diet. One can be self-disciplined and still flexible enough to

make exceptions on those rare occasions when large benefits would be lost

if one kept to the regimens one followed. Likewise, then, one could have the

dispositions of a cooperator and still be flexible enough to cheat on those

rare occasions when cheating brought large benefits without more than a

negligible risk of discovery.

Perhaps, though, the dispositions of a cooperator are different from

those of self-discipline. Perhaps no one who saw the importance of cooper-

ation could at the same time be willing to cheat on rare occasions. This

possibility would make sense if the importance of cooperation were differ-

ent from that of a diet or a budget, if for instance it were important to the

person for reasons other than its contribution to her happiness. But obvi-

ously this possibility is not one to which a defender of Hobbes�s answer

could appeal. On Hobbes�s answer the cooperation of others is important to

people because without it their lives would be miserable and short. So the

analogy of the dispositions of a cooperator to those of self-discipline is apt.

These are the dispositions that anyone clearheaded about the importance of

cooperation would cultivate in himself.

The more sophisticated version of Hobbes�s answer thus appears to be

no sounder than the simpler one. It depends on the assumption that the

paradigms of the dispositions of character whose manifestation in one�s

dealings with others elicits their trust are the traditional virtues like

justice and honesty, and this assumption appears to be unwarranted. We

have, it seems, no more reason to think that a person best promotes his

happiness by cultivating in himself these traditional virtues than we have

to think that he best promotes it by strictly following the standards gen-

eral adherence to which enables people to get along and work together.

The Hobbesian program of grounding these standards in the fundamental

principle of egoism is equivalent to a program of showing that the funda-

mental principle endorses cultivating in oneself the traditional virtues

like justice and honesty. One makes no progress, then, in trying to shore

up the former by appealing to the latter. The Hobbesian program, there-

fore, in the absence of a better answer than Hobbes�s to the challenge

posed by the free-rider, falls short of vindicating these standards as the

standards of morality.

Egoism 51

9. Troubles with the Hobbesian program�s scope

A second criticism of the program concerns its scope. Hobbes restricted his

derivations of natural law to standards of conduct general adherence to

which enabled people to get along with each other and work together. At

the same time, he acknowledged other standards that he believed were

derivable from natural facts about human existence and what a human

being needs to live a tolerable life. These other standards concerned per-

sonal safety and health. Examples are standards of moderation in drinking

intoxicating beverages and eating rich food. There are, however, still other

commonly observed standards of conduct, and while some of these, like

manners and etiquette, are essentially local and conventional, others, like

standards of kindness toward nonhuman animals, appear not to be. An

ethical theory can ignore the former, since being essentially local and

conventional they do not represent standards whose authority in practical

thought one could consider as being backed by reason rather than custom.

But it cannot ignore the latter. If it excludes them from morality without

explanation or a showing that they too, despite appearances, are essentially

local and conventional, it is open to the criticism of being too restrictive.

And egoism is open to just this criticism. Because it recognizes as standards

of morality only standards conformity to which promotes the actor�s own

happiness, it excludes from morality standards that would otherwise seem

perfectly good candidates for being standards whose authority in practical

thought is backed by reason rather than custom.

One example will suffice. Consider the rule proscribing cruelty to nonhu-

man animals. This rule, unlike the one that proscribes cruelty to human

beings, is not a standard one could hope to ground in the fundamental

principle of egoism by applying the argument of the Hobbesian program.

That argument applies to the rule proscribing cruelty to human beings since

a personwho is cruel to other humanbeings is not someone peoplewill trust.

But a person who is cruel to nonhuman animals is not similarly distrusted.

One can speculate on the reasonswhy.One reason, presumably, is that, being

less likely to identify with beasts than with our fellow humans, we are less

likely to concern ourselves with being possible victims of a person�s cruelty

when it is inflicted on a nonhuman animal than when it is inflicted on

another humanbeing. Butwhatever the reasons, the fact is plain. Fox hunters


who are indifferent to the pain and suffering of their quarry, cosmetics

researchers who subject the laboratory animals on which they test new

products to unconscionable physical torment, breeders of gamecocks who

stage cock fights for the fun of watching two birds fitted with metal spurs

bloody each other till one succumbs do not find themselves losing the trust

of others because of these activities. Egoism, then, cannot recur to the

Hobbesian program to establish the rule proscribing cruelty to nonhuman

animals as a standard of morality. Nor does there appear to be any other

secure basis on which the theory could establish this. Cruelty to nonhuman

animals, it would seem, is not necessarily contrary to achieving happiness.

Egoism, for this reason, appears too restrictive in what standards it can

recognize as part of morality.

10. Thrasymachus� challenge again

Over and above the particular problems that beset the Hobbesian program,

there remains to be reckoned with the challenge to the very project of

ethical theory that Thrasymachus� moral skepticism represents. It chal-

lenges the project of ethical theory by denying the existence of the stand-

ards that ethical theory takes as its object, the standards of morality. It

denies, that is, that there are any universal standards of conduct whose

authority in practical thought derives from reason rather than custom. This

challenge, as we noted earlier, is particularly sharp in the case of egoism,

since Thrasymachus, too, thinks that one�s own happiness or well-being is

the highest good one can achieve. But because he also thinks that some

people are superior to others, he rejects the idea of a set of universal stand-

ards of conduct whose authority every human being, through an exercise of

reason, would recognize. Thus, his challenge differs importantly from the

challenge to egoism posed by the free-rider. Anyone could be a successful

free-rider. Anyone could evade standards that others are following and

thereby reap the benefits of their cooperative behavior while avoiding the

costs of cooperation. But not just anyone, on Thrasymachus� view, can

successfully flout the standards of justice in the imperious way he applauds.

Rather, in his view, such conduct is reserved for the few whose superiority

in power enables them to practice such injustice with impunity. What is

more, the question of the risk of their being discovered is not an issue for

him. As far as he is concerned, they can practice injustice openly, for

Egoism 53

overawing power and not stealth is the key to their success. Answering the

challenge his skepticism represents, then, does not require making the risk

of discovery real. It requires, instead, arguing against Thrasymachus� excep-

tionalism about those whom he regards as superior human beings.

The argument answering the challenge becomes evident once one puts

Thrasymachus� exceptionalism in the terms of the Hobbesian program.

Accordingly, Thrasymachus� view is that some people are so powerful that

they can achieve happiness even if no one trusts them. In other words, none

of them needs to manifest the dispositions of a cooperator in his dealings

with others to get them to cooperate with him. Fear of his power alone is

sufficient, Thrasymachus must have supposed, for securing their coopera-

tion. And hemust have further supposed that because of this tyrant�s power,

the tyrant has nothing to fear from others and so cooperation with them is

unnecessary. Clearly, these two suppositions amount to a denial of the fact of

equality on which Hobbes�s derivations are conditioned. Thus, the argument

answering Thrasymachus� challenge consists in showing that these supposi-

tions are inconsistentwith this fact. In short, it is that no one is so powerful as

to be invulnerable to lethal attack and mortal injury. Because human beings

are all equal in this respect, even powerful tyrants have something to fear

from others. For them to practice injustice on the grand scale Thrasymachus

imagined is for them to act inways that, however powerful they are, increase

the risk of their being the victimof lethal attack, and an increase in this risk is

not worth the benefits of the injustices they practice. These tyrants, too, have

reason to cultivate in themselves the dispositions of a cooperator.

An admirer of Thrasymachus� position could dispute this, of course. But

to rebut it he would have to do more than give examples of happy tyrants

who greatly enriched themselves at the expense of their subjects, lived long

lives, and died of natural causes. For the same point applies here as the one

Hobbes made in answering the challenge posed by the free-rider. Because

success alone may be due to luck, one cannot simply just give examples of

successful tyrants to rebut the argument. Thus to rebut it our admirer of

Thrasymachus would also have to show that their escaping from a prema-

ture and violent death was to be expected from the provisions for safety

they made, that owing to these provisions they could count on being safe

from such violence despite their odious conduct. And it is far from certain

that he can show this. Citing the tyrants� power in terms, say, of the

weaponry and soldiery at their command or the wealth they could draw


on would not show it. Such power alone does not provide the necessary

assurance. There is also the tyrants� need of the loyalty of their lieutenants,

and for any tyrant who practices injustice on a grand scale retaining this

loyalty over the long run is surely a chancy thing. Here at any rate is the

weak point in the admirer�s rebuttal. Even powerful tyrants have something

to fear, and the less attention they give to manifesting the dispositions of a

cooperator the more their lives are ruled by fear. In view of the fact of

equality on which Hobbes conditioned his derivations, a defender of ego-

ism, then, has a reasonable, if not conclusive, answer to the challenge that

Thrasymachus� skepticism represents.

When Glaucon restates Thrasymachus� position, he replaces the tyrant

who practices injustice with impunity with a shepherd-turned-criminal,

Gyges, whose ability to commit crimes with impunity is due to his having

the power, thanks to a magical ring, to make himself invisible at will. The

switch is not inconsequential. Thrasymachus bases his moral skepticism on

his demographic distinction between the powerful and the weak, and with

the switch of examples this distinction is lost.What replaces it is a distinction

between people who have magical powers and the rest of us, and this new

distinction cannot serve as a basis for moral skepticism. The upshot is that

Glaucon, by switching examples, replaces the challenge that Thrasymachus�

skepticism represents with a different one. It is the one posed by the free-

rider. For Gyges is the ultimate free-rider. Yet Plato does not mean this

example to be a disproof of egoism. His point is not to show that, on the

definition of right action that egoism endorses, criminal action is the right

thing to do if you have a magical ring that enables you to make yourself

invisible at will. His point, rather, is to question the relation that egoism

supposes right action bears to the highest good. When the highest good is

identified with a life filled with pleasure unspoiled by pain, right action is

understood as instrumental to producing pleasure and escaping from pain.

Plato�s view is that the relation right action bears to the highest good is

different. It is such that one could not even imagine someone�s achieving

the highest good without living according to the standards of right action. To

hold this view, however, requires that one make right action inseparable in

conception from the highest good and therefore reject hedonism as the

theory of human well-being. Plato did both in constructing a theory on

which the development and exercise of the traditional virtues is integral to

realizing the highest good. It is this theory that we shall take up next.

Egoism 55

3 Eudaimonism

1. Egoism v. eudaimonism

Eudaimonism was the dominant theory in ancient Greek ethics. The name

derives from the Greek word �eudaimonia,� which is often translated

�happiness� but is sometimes translated �flourishing.� Many scholars in

fact prefer the latter translation because they believe it better captures the

concern of the ancient Greeks with the idea of living well. This preference

suggests that a useful way of distinguishing between eudaimonism and

egoism is to observe, when formulating their fundamental principles, the

distinction between well-being and happiness that we drew in the last

chapter. Accordingly, the fundamental principle of eudaimonism is that

the highest good for each person is his or her well-being; the fundamental

principle of egoism remains, as before, that the highest good for a person is

his or her happiness. Admittedly, this way of distinguishing between the

two theories would be theoretically pointless if the determinants of how

happy a person was were the same as the determinants of how high a level

of well-being the person had achieved. Thus, in particular, when hedonism

is the favored theory of well-being, this way of distinguishing between

eudaimonism and egoism comes to nothing. It fails in this case to capture

any real difference between them. For when hedonism is the favored theory

of well-being, determinations of how happy a person is exactly match the

determinations of how high a level of well-being a person has achieved.

Matters are different, however, when perfectionism is the favored theory

of well-being. In this case, one can capture a real difference between eudai-

monism and egoism by distinguishing one from the other according as its

fundamental principle takes a person�s well-being or his happiness to be the

highest good. For on a perfectionist conception of well-being, determina-

tions of how happy a person is do not match the determinations of how

high a level of well-being a person has achieved. The determinants are not

56

the same, as our example in the last chapter of the aspiring violinist shows.

Hence, when one takes perfectionism as the favored theory of well-being,

this way of distinguishing between eudaimonism and egoism has theoret-

ical significance, and accordingly one can treat the former as a separate

object of study. Plato�s ethics is its original form. To answer the challenge

put to Socrates by Glaucon and Adeimantus, Plato argued for perfectionism

as an advance over hedonism in the theory of well-being. He thus, in effect,

showed us how to fix eudaimonism as a distinct ethical theory from egoism.

Nonetheless, you might still wonder whether it made sense to draw this

distinction. Youmightwonder, specifically, whether itmade sense to define

�egoism� so narrowly that it applied only to a theory whose fundamental

principle took a person�s happiness, rather than his or her well-being, as the

highest good for that person. Granted, there is a real difference between

eudaimonism and egoism on this way of distinguishing between them, but

you might still wonder whether this difference was too small to warrant

treating the former as a separate theory. Why isn�t a theory that identifies a

person�s highest goodwith hiswell-being also egoistic? It would be, after all,

as we noted earlier, if hedonism were the correct theory of well-being. So

why is it not also egoistic if perfectionism is the correct theory of

well-being?

The answer comes from the pivotal point in Plato�s rejection of hedon-

ism. Plato rejects egoism and turns to perfectionism when he hits upon the

idea of a noninstrumental relation between right action and the highest

good. If eudaimonism could not accommodate this idea, then we would

learn nothing from treating eudaimonism as a distinct theory from egoism.

Doing so would yield no significant results apart from those that were

directly due to the difference between perfectionism and hedonism. In

that case, we might just as well define �egoism� more broadly to include

eudaimonism and then distinguish between the latter�s hedonistic and

perfectionistic versions. But because eudaimonism can accommodate

Plato�s idea, it is possible to elaborate it in a way that represents right action

as action worth doing for its own sake and, indeed, as a constituent of the

highest good. For this reason it is appropriate to treat it as a distinct theory

from egoism and therefore to define �egoism� narrowly so as to exclude it.

Specifically, it is appropriate to do so because our idea of egoism is that of

a theory whose conception of the highest good is consistent with the

possibility of someone�s realizing it by being a free-rider. Accordingly, a

Eudaimonism 57

theory onwhich right action is a constituent of the highest good does not fit

this idea. The reason is clear. If right action is a constituent of the highest

good, no one can realize the highest good by being a free-rider. The incor-

poration of Plato�s idea into a perfectionistic account of well-being thus

yields a form of eudaimonism that, because it avoids the challenge free-

riders pose for egoism, offers theoretical possibilities beyond the scope of

egoism. It is no surprise, then, that this specific form of eudaimonism,

owing to its being, unlike the other forms, differentiable from egoism, has

come to be identified with the general theory.

2. The Platonic form of eudaimonism

Plato embedded his idea in the problem that Glaucon and Adeimantus put

to Socrates when they restated Thrasymachus� position. The problemwas to

show conclusively that a person who practices justice gains, from acting

justly, benefits that he could not otherwise get and that surpass any that he

could get by appearing to be just without really being so. Conventional

wisdom, Glaucon pointed out, does not commend practicing justice as a

way of acting towards others that is salutary and worthwhile in its own

right. Rather it says that the reason one practices justice is to gain a repu-

tation for being a just person, for such a reputation secures the trust of

others and thereby brings opportunities from which people known to be

unjust are typically excluded. Yet if the benefits of such a reputation were

the sole reason a person had for being just, then he could gain themmerely

by appearing to be just when he really wasn�t, provided, of course, that his

appearance was convincing. Glaucon makes this point with his example of

Gyges� ring. The example, moreover, not only illustrates this specific point

but alsomakes clear themore general one that appearing to be just can be at

least as profitable as really being just so long as practicing justice brings a

person no benefit that he cannot gain in some other way. In other words,

appearing to be just can be at least as profitable as really being just so long as

practicing justice is never more than a means to the benefits it brings.

To solve the problem, then, that Glaucon and Adeimantus put to Socrates

requires showing that practicing justice is constitutive of living well and not

merely a means to it. That is, it requires showing that practicing justice is an

essential ingredient in the achievement of a durable state ofwell-being. Plato,

one could say, set a problem for himself whose solution required a theory on


which a person�s well-being included among its integral parts his being just

and thus having the disposition to act justly. Such a state, on any theory that

solved this problem, could not be a state that the exercise of that disposition

was merely instrumental in producing. Plato thus set terms for a form of

eudaimonism whose theoretical possibilities outstripped those of egoism.

Generally speaking, every form of eudaimonism affirms the ideal of the

wise pursuit of well-being. If one takes hedonism as the favored theory of

well-being, then one obtains a form of eudaimonism that affirms the ideal

of the wise pursuit of pleasure and freedom from pain. This form, as we

noted above, is indistinguishable from egoism when one makes the same

assumption about hedonism�s being the favored theory of well-being.

Correspondingly, then, egoism, on this assumption, affirms the very same

ideal. By contrast, if one takes perfectionism as the favored theory of well-

being, then one obtains a form of eudaimonism that affirms the ideal of the

wise pursuit of excellence in the service of ends whose achievement makes

life worth living. Or put more simply, since wisdom is itself an excellence

and since it entails the adoption of ends whose achievement makes life

worth living, the ideal this form of eudaimonism affirms is that of the wise

pursuit of excellence. This is plainly a different ideal from the one implicit

in egoism. While one can pursue happiness and excellence at the same

time, achieving the one is not the same as achieving the other. Our example

of the aspiring violinist who is uncertain about how high to set her sights

makes this clear.

Moreover, if one takes the Platonic form of eudaimonism as the best

representative of those forms that assume perfectionism as the favored

theory of well-being, then the ideal the form affirms is that of the wise

pursuit of intellectual and moral excellence. For on the Platonic form

excellence consists in wisdom in the choice of ends as well as means to

achieving those ends and also in such dispositions as justice, honesty, and

courage, which keep one�s choices from straying from right action. This

ideal differs evenmore starkly from the ideal that egoism affirms. Because a

person�s happiness can be more reliably secured if she does not aspire

always to do what is most just or courageous, it takes luck to achieve

both. Someone, aware of this fact, who aspires to intellectual and moral

excellence gives up any aspiration to happiness. The study of eudaimonism,

then, when it focuses on the Platonic form of the theory, departs signifi-

cantly from the study of egoism.

Eudaimonism 59

Whether the Platonic form is the soundest form is, to be sure, another

matter. It is to this question that we now turn.

3. Perfectionist objections to hedonism

The choice between different forms of eudaimonism depends, in the first

place, onwhether to favor hedonismor perfectionism as the sounder theory

of well-being. The question comes down to whether the highest good con-

sists in one�s experiencing maximal pleasure and freedom from pain in life

or in one�s achieving the greatest degree of excellence in life. It comes down

to whether there is more reason to affirm pleasure or excellence as the

ultimate end of one�s actions. In other words, would a life that is filled with

asmuch pleasure and as little pain as possible be the best life one could live,

or would the best life be one in which one achieved excellence, to the

greatest degree possible, in pursuits whose undertaking make life worth

living? Pleasure is no doubt among the first things, if not the very first thing,

a person considers to be the highest good. Likewise hedonism is among the

oldest theories of well-being, if not the oldest. The theory has a prominent

place, for instance, in the Protagoras, one of Plato�s early dialogues, and it

was also advocated in Plato�s time by other followers of Socrates.1

Excellence, by contrast, does not so quickly recommend itself as the highest

good, especially since many of the excellences one first thinks of are gifts

and skills, like agility and intelligence, efficiency and technical facility, that

seem worth having, not for their own sake, but because of the advantages

having them brings. Nevertheless, perfectionism, too, goes back to Plato. It

arises out of his critique of hedonism. Hedonismmay be the oldest theory of

well-being, but opposition to it is virtually as old, and from this opposition,

thanks to Plato�s genius, a powerful alternative emerged.

The main objection to hedonism is that to regard pleasure and freedom

from pain as the highest good for a human being is to regard the life of

human beings as no different in kind from the life of other animals capable

of experiencing pleasure and pain. J. S. Mill, in his discussion of hedonism,

put the objection directly when he wrote:

Now such a theory of life excites in many minds, and among them in some

of the most estimable in feeling and purpose, inveterate dislike. To suppose

1 Aristippus of Cyrene (435–356 BC).


that life has (as they express it) no higher end than pleasure – no better and

nobler object of desire and pursuit – they designate as utterly mean and

groveling; as a doctrine worthy only of swine, to whom the followers of

Epicurus were, at a very early period, contemptuously likened; and modern

holders of the doctrine are occasionally made the subject of equally polite

comparisons by its German, French, and English assailants.2

Mill, it should be said, wrote as a defender of hedonism and not as a critic.

His statement of the objection occurs as a prelude to his answering it. But

before we consider his peculiar answer, let us press the objection further

and in a direction that more clearly encompasses Plato�s alternative.

Experiences of pleasure may be more or less intense and more or less

long-lasting. When one considers examples, common ones include gratifi-

cation of a strong appetite and the kind of feelings that go with relief from

physical stress or discomfort, the kind of feelings that relaxing in a comfort-

able chair or a warm bath produces. Such pleasures are not unique to

human life. Cats, too, can have strong appetites and experience intense

pleasure from their gratification, and though they do not like warm baths,

they do enjoy curling up in comfortable chairs. Similar observations apply

to cows, pigs, moose, otters, and rats. Hence, so the objection goes, if the

best life a human being can live is a life that is filled with pleasures of

varying intensity and duration and free of all but perhaps the most transi-

tory and bearable pains, then that person�s highest good would seem to be

no different from a cat�s highest good or a cow�s. By taking pleasure and

freedom from pain as the sole elements of well-being, hedonism would

seem to have reduced humanwell-being to animal well-being and therefore

to have denied that it consists in anything special to human existence.

Except for the possibility that our distinctively human powers and capabil-

ities enable us to experiencemore intense pleasures and longer-lasting ones

than other animals, hedonism offers us no reason to prefer being human to

being a cow. Yet a bovine life, however much pleasure and little pain it

might contain, surely lacks many of things that make human life worth

living. Contrary to hedonism, then, so it appears, there is more to human

well-being than pleasure and freedom from pain.

The same conclusion, moreover, appears to follow from a further objec-

tion. Thus, on the account of human well-being that hedonism endorses, a


Eudaimonism 61

person could not only live a good life by spending it gratifying his appetites

and relaxing in stress-free circumstances, but his life could also be made

better by his enlarging and then gratifying his appetites and by his finding

increasingly snug circumstances in which to relax. In the end, it would

seem, the best life for a human being, according to hedonism, could be

realized by a person�s gorging himself on rich foods, downing large quanti-

ties of thirst-quenching beverages, copulating in as many ways and with as

many partners as his enlarged sexual cravings demand, and spending the in-

between times bathing, sleeping, and lounging around on plush sofas or in

gently swaying hammocks while peacefully listening to soothing music.

Such an indulgent life, however, is not the picture of human well-being, for

the trajectory it represents is that of descent into a kind ofmindless bliss. Its

terminus would be a life in which one became solely a receptor of pleasure-

giving and nutritive substances, perhaps delivered intravenously, and in

which the days were passed entirely listlessly in a drug-induced state of

continuous pleasure. Call this the life of total passivity. Itmay not represent,

on hedonism�s account of well-being, realization of the highest good. It is

possible that the highest good is realized in a life in which some pleasures

are not drug-induced and in which the pursuit of pleasure is not entirely

passive. But hedonism�s account of well-being cannot rule out a priori the

possibility that the life of total passivity does represent the best life for a

human being. For this reason the account appears misconceived.

Specifically, it appears to yield an inapt conception of human well-being.

Plato, too, believed that hedonism�s account of well-being would be seen

as misconceived when its endorsement of a life consisting of constant

gratification of larger and larger appetites was pushed to the limit. Such a

life, according to Plato, was destined to end in madness. As a person�s

appetites grew ever larger, Plato argued, his or her personality would

become progressively disordered. The end results were a derangement of

the mind, like that experienced in dreams, and a breakdown of self-control,

like that exemplified by the most impulse-driven maniacs. To make this

argument, though, Plato had to draw on his own conception of the human

personality and the perfectionist theory of well-being it supports. Hence, as

a criticism of hedonism, his argument is too closely tied to this rival account

of well-being to be telling. Criticisms that are independent of rival accounts

are more effective. Nevertheless, Plato�s criticism is important in the way it

sets the perfectionist account to which it is tied against hedonism. The


opposition between the two illuminates the traditional themes of the

perfectionist tradition that he began.

Thus, according to Plato, one achieves the highest good through the

development of a well-ordered personality. One achieves it, that is, by

cultivating and preserving in oneself the best arrangement of the various

traits and powers that constitute one�s personality. In so doing one realizes

the best life one can live as a human being. In such a life, the constituent

parts of one�s personality work together harmoniously. These are, for Plato,

following his famous tripartite division of the soul, reason, spirit, and

appetite, and harmonious cooperation among them results when reason,

with the help of the emotional energy Plato identified as spirit, regulates

and constrains the appetites. Plato argued for this thesis from an analogy he

drew between the constitution of a man�s or woman�s personality and the

constitution of a city, a city being the primary political society or state in

Plato�s time. Accordingly, Plato observed, just as in the best-governed city

the wisest men and women, with the help of an auxiliary martial force, rule

over a compliant and prosperous proletariat, so too in the best-governed

soul, reason, the seat of intelligence and knowledge, with the help of spirit,

regulates and constrains the appetites. And from a similar comparison

between poorly governed cities and dysfunctional individuals, Plato

inferred that disharmony in the soul resulted when a person�s appetites,

aided by spirit, became the ruling forces in his personality. The ascendancy

of the appetites corresponds to a decline in the influence of reason, and this

decline brings a loss of coherent thought and a breakdown in self-control.

Hedonism, therefore, Plato maintained, insofar as it takes gratification of

the appetites as the chief source of the highest good, advances a specious

account of human well-being.

4. Epicurus� answer

A defender of hedonism has no trouble answering Plato�s criticism directly.

Since madness is bound to bring more pain and less pleasure than sanity,

the best human life, on the hedonist account, cannot tend towardmadness.

Hedonism, too, counsels that one discipline one�s appetites. The grounds of

this advice are that the pleasures of gratifying appetites without restraint

are impure and their impurity, which is to say, the painful consequences of

such gratification, reduces the value of the experiences to something less

Eudaimonism 63

than its cost, as measured by the amount of pain one later endures. Severe

stomach cramps that follow overeating is an obvious example. A defender

of hedonism, then, canmaintain, in answer to Plato, that theway to achieve

the highest good, on the hedonist account of it, is to balance the value of the

available pleasures against their cost and to seek those that either have the

least impurity or exceed in value to a greater extent than others the cost in

pain that experiencing them entails. This answer represents the wise pur-

suit of pleasure and freedom from pain, which is the ideal implicit in

hedonistic eudaimonism, and in practice this pursuitmay differmarginally,

if at all, from the pursuit of a well-ordered personality that Plato recom-

mended. Indeed, the leading defender of hedonism among ancient Greek

philosophers, Epicurus, maintained as a central theme of his ethics that the

wise pursuit of pleasure and freedom from pain entailed avoiding exciting

pleasures because of their impurities. One should seek, instead, according

to Epicurus, quieter, contemplative pleasures, and the life of contemplative

pleasure that he recommended closely matches the life of philosophy that

Plato extolled.3

Epicurus� defense of hedonism thus answers Plato�s criticism. It does so

by showing that in practice the wise pursuit of pleasure does not lead to a

life spent feeding one�s appetites. Instead, it leads to a life spent in the

cultivation of one�s intellectual powers and their contemplative exercise. In

practice, then, on this Epicurean defense, hedonism promotes a kind of life

organized around distinctively human powers and aspirations, and such a

life is not open to Plato�s criticism. Yet showing that hedonism in practice

promotes such a life as the best life for human beings is not the same thing

as showing that the theory can conceive of no other life as being better. In

particular, it is not the same thing as showing that, as a matter of principle,

hedonism rejects the life of total passivity as the best life for human beings

or denies that the best human lifemay be no better than the best bovine life.

Consequently, a defense of hedonism, like Epicurus�, that appeals to the

consequences of hedonistic practice does not shield the theory from

criticisms of its conception of human well-being, for the conception may

still be in error even if, owing to the actual contingencies of human life, the

theory works out well in practice. Plato maintained that it failed disas-

trously in practice, that it led to wantonness and ultimately to madness,

3 See Epicurus, �Principle Doctrines.�


and a defender of hedonism can answer Plato�s charge, as Epicurus did. But

to answer criticisms that target hedonism�s conception of human well-

being, a different kind of defense is needed.

5. Mill�s defense of hedonism

Mill�s defense is of this kind. Hedonism, he declared, when properly under-

stood, does not conceive of human well-being as comparable to the well-

being of other animals. Rather it elevates human well-being above that of

every other animal. To support this view, Mill first distinguished the kinds

of pleasure that only human beings are capable of experiencing and then

argued for the superiority of these to the kinds the experience of which is

common to both humans and other animals. His argument for the superi-

ority of the former depends on a novel thesis he put forward about the

factors determining the intrinsic value of an experience of pleasure. On the

standard account of these factors, they are all quantitative: one estimates a

pleasure�s intrinsic value by gauging its intensity and duration, which are,

in principle,measurable in units of force and time.Mill�s thesis, by contrast,

was that the quality of a pleasure, as well as these quantitative factors,

determined its intrinsic value. �It would be absurd,� Mill wrote, �that

while in estimating all other things quality is considered as well as quantity,

the estimation of pleasure should be supposed to depend on quantity

alone.�4 And he further held that the experiences of pleasure of which

only human beings were capable so far exceeded in quality the experiences

of pleasure common to both humans and other animals as to make consid-

erations of intensity and duration inconsequential when one was esti-

mating their intrinsic value. As he put the point himself, they have �a

superiority in quality so far outweighing quantity as to render it, in compar-

ison, of small account.�5 From this point and his novel thesis, hedonism�s

elevation of human well-being above that of every other animal follows

immediately.

It follows, too, that hedonism does not require a conception of human

well-being onwhich a life spent gratifying one�s appetites could, in principle,

be the best life for human beings. Quite the contrary, Mill maintained,

hedonism could avail itself of a conception of human well-being on which

4 Mill, Utilitarianism, ch. 2, par. 4. 5 Ibid., ch. 2, par. 5.

Eudaimonism 65

such a life was necessarily worse than a life in which one found pleasure in

the exercise of such distinctively human powers as reason, imagination, and

moral sentiment. Such pleasure, Mill argued, had greater intrinsic value than

the pleasure of gratifying one�s appetites. Its intrinsic value was greater

because the pleasure was qualitatively superior to that of gratifying one�s

appetites and because the difference in quality was sufficiently great as to

make considerations of quantity of little account. Hence, Mill concluded, a

life enriched by the pleasure one gets from successfully engaging in activities

that draw on one�s distinctively human powers was necessarily better than

one in which the pleasure of gratifying one�s appetites predominates.

In effect, then, Mill�s defense of hedonism is an attempt to disarm the

criticism of hedonism that perfectionism inspires. His revision of hedonism

to include quality as one of the factors that determine the intrinsic value of

an experience of pleasure enables the theory to grade the kinds of pleasure

human beings are capable of experiencing independently of their intensity

and duration and to place such value on experiences of high-grade pleasure

in comparison with experiences of low-grade pleasure as to make the con-

tribution to human well-being of the latter of little consequence, regardless

of their intensity and duration. Mill�s defense, if successful, could thus save

hedonism from having, in principle, to allow that a life in which experi-

ences of low-grade pleasure predominated could be the best life for human

beings. Indeed if, as Mill maintained, high-grade pleasures were the pleas-

ures one derived from exercising the abilities and strengths that counted as

human excellences, then the defense would seem to allow hedonism to

endorse as the best life for human beings the very same life as the one

perfectionism endorsed. Hence, if the defense were successful, the opposi-

tion between hedonism and perfectionism would become merely abstract

and the issue between them moot.

To be sure, whether the defense is successful depends on the soundness of

Mill�s thesis that quality, along with intensity and duration, is a factor deter-

mining the intrinsic value of experiences of pleasure. It depends, too, on the

truth of his identifying high-grade pleasures with the pleasures one derives

from exercising distinctively human powers and low-grade pleasures with

the gratification of animal appetite. Why, after all, should one think that the

one kind of pleasure is qualitatively any better, as a kind of pleasure, than the

other? Mill answered this question directly. The ultimate test, he declared, of

whether one kind of pleasure is qualitatively better than another is the


preferences of those who have experienced both kinds and are capable of

appreciating either. If such experts in pleasure clearly prefer one kind over

the other, then, according toMill, their preferences prove beyond doubt that

the kind they prefer is qualitatively better. �Of two pleasures,� he wrote, �if

there be one to which all or almost all who have experience of both give a

decided preference, irrespective of any feeling of moral obligation to prefer

it, that is the more desirable pleasure.�6 And he went on to state that people

who have experienced and can appreciate both the kind of pleasure that

humans derive from the exercise of their higher human powers and the kind

that comes from gratification of their animal appetites have a clear prefer-

ence for the former over the latter. In Mill�s words, �[I]t is an unquestionable

fact that those who are equally acquainted with and equally capable of

appreciating and enjoying both do give a most marked preference to the

manner of existence which employs their higher faculties.�7

How convincing is this answer? To be convinced one must first agree

with Mill about the ultimate test of whether one kind of pleasure is

superior in quality to another, and agreement about this test is hardly a

given. To the contrary, some of Mill�s sharpest critics – notably G. E. Moore

(1873–1958) – have rejected it.8 Suppose Mill were right, Moore argued,

and those who had experienced and could appreciate each of two kinds of

pleasure had a marked preference for one kind over the other; then these

experts would have tofind the kind they preferredmore desirable than the

other. But for one kind of pleasure to be more desirable than another it

must either be more pleasurable, in which case the quantity of the pleas-

ure, not the quality, determines its desirability, or have some property

besides its pleasantness that positively contributes to its being desirable

and that the other kind lacks, in which case pleasure and freedom from

pain are not the only determinants of desirability. Hence, Mill�s viewmust

fail since, on his view, neither of these alternatives is possible. The first

requires giving up the thesis that quality as well quantity is a factor

determining the intrinsic value of pleasure. The second requires giving

up hedonism. Moore�s objection, though, falls short of being a refutation.

One can resist it. (Couldn�t one pleasure�s being more pleasurable than

another be a qualitative and not a quantitative difference? Couldn�t it be

6 Ibid. 7 Ibid., ch. 2, par. 6.8 See G.E. Moore, Principia Ethica (Cambridge: Cambridge University Press, 1903), pp. 77–81.

Eudaimonism 67

like one argument�s being more subtle than another?) Still, even if one

agrees withMill and accepts his test, one has reason not to be convinced by

his answer. For the answer rests on an unwarranted assumption.

The assumption is implicit in the test itself. To take the preferences of

those who have experienced and can appreciate each of two kinds of

pleasure as the ultimate test of whether one kind is superior in quality to

the other is to assume that the object of these experts� preferences is

pleasure and not, say, the things or activities from which they derive

pleasure. An analogy should make this clear. We regard the preferences of

those who know wines as proof of which merlots, say, are superior because

we think such experts are better able, in virtue of their experience and

knowledge, to discriminate among different grades of wine. Likewise, we

regard the preferences of those who know the voice as proof of which

sopranos have superior voices because we think such experts are better

able, in virtue of their experience and knowledge, to discriminate among

different qualities of voice. In either case, we take the preferences as reliable

indicators of expert judgment about the wine�s or the voice�s quality, and

there is no better proof of the quality of a wine or voice than such judgment.

Yet the preferences count as proof only because their objects are the taste of

thewine, rather than, say, its viscosity, or the sound of the voice rather than

the movements of the mouth out of which it comes. If their objects were

different, then they would not reliably indicate expert judgment about

which wine or voice was superior in quality. The same, then, must be true

of the preferences to whichMill appeals as proof of the superior quality of a

pleasure. They, too, would not reliably indicate expert judgment about

which kind of pleasure was superior if their object were not that pleasure

but the things or activities that are its source. They would not, therefore,

count as proof of the superior quality of pleasure of that kind.

Mill, then, must be assuming that the object of the preferences to which

he appeals is the pleasure of exercising higher human faculties and not the

faculties themselves or the activities that employ them. But on what basis

can he make this assumption? He thinks it is �an unquestionable fact,� as

we saw, that people who know both this kind of pleasure and the kind that

comes from gratification of animal appetite �give amostmarked preference

for the manner of existence that employs the higher faculties,�9 yet such



people could have this preference without also preferring the pleasure they

get from exercising those faculties. Indeed, Mill appears to accept this point

when he writes:

Few human creatures would consent to be changed into any of the lower

animals for a promise of the fullest allowance of a beast�s pleasures; no

intelligent human being would consent to be a fool, no instructed person

would be an ignoramus, no person of feeling and conscience would be selfish

and base, even though they should be persuaded that the fool, the dunce, or

the rascal is better satisfied with his lot than they are with theirs. They

would not resign what they possess more than he for the most complete

satisfaction of all the desires which they have in common with him.10

And again, when he declares, several sentences later:

It is better to be a human being dissatisfied than a pig satisfied; better to be

Socrates dissatisfied than a fool satisfied. And if the fool, or the pig, are of a

different opinion, it is because they only know their side of the question. The

other party to the comparison knows both sides.11

These remarks, far from supporting his assumption, accent our uncer-

tainty about it. The preferences they describe appear to be preferences

for a life that employs the higher faculties despite its being less pleasura-

ble than a life employing only the lower ones. They appear, that is, to fit a

perfectionistic outlook rather than a hedonistic one. So what could

Mill�s reason be for thinking his assumption was consistent with this

description?

The answer can only be Mill�s belief that every preference is at bottom a

preference for pleasure. It can only be, in other words, the doctrine of

psychological hedonism, to which Mill, as we noted in the last chapter,

firmly adhered. Plainly, if this doctrine were true, then Socrates� prefer-

ence for the philosophical life, despite its being less satisfying than a

fool�s, would have to be a preference for the pleasures distinctive of that

life. But it is equally plain that if the doctrine were false, there would be no

reason to think Socrates� preference was a preference for pleasure. On its

face, after all, it is a preference for living a life befitting a rational being,

even when such a life brings less pleasure than one suited to fools. So to

justify taking it as at bottom a preference for pleasure, one needs to assert

10 Ibid. 11 Ibid.

Eudaimonism 69

some basic principle of psychology according to which the real object of

such preferences is different from their manifest object. And in this case

the principle would have to be the doctrine of psychological hedonism.

More generally, then, one must assert this principle to justify the assump-

tion that the object of the preferences for a life requiring the exercise of

higher human faculties that all whom Mill cites as experts in the matter

display is the pleasure of exercising those faculties. In other words, the

assumption implicit in Mill�s test for the qualitative superiority of one

kind of pleasure over another is based on the doctrine of psychological

hedonism. Working backwards, then, from this result one can see that

Mill�s defense of hedonism is itself built on his presumption of this doc-

trine. It is built on his fundamental belief that pleasure is the ultimate

object of every preference.

One can see, too, how, if the doctrine of psychological hedonism were

true, Mill�s defense would succeed in disarming the criticisms of hedon-

ism that perfectionism inspires. The problem, however, is that the doc-

trine is not obviously true and is, in fact, open to serious doubt. Certainly

perfectionists would reject it as a principle of psychology. Echoing

Butler, they would insist that the pursuit of excellence is the pursuit of

excellence, and not another thing. Mill�s defense of hedonism, therefore,

succeeds only by presupposing a principle that its opponents would

reject, and not unreasonably. Its success is, in a word, hollow. Nor are

perfectionists the only ones who regard the doctrine as false. The diffi-

culty of sustaining it that Butler�s criticisms show has persuaded many

other thinkers as well that it is false. Nonetheless, it remains a perennial

of moral philosophy, which continues to attract new believers while

retaining its stalwarts, and they may still look hopefully to Mill�s defense

for a way to escape the main objections to hedonism�s conception of

human well-being. Yet given the difficulty of sustaining the doctrine,

they might just as well plump directly for that conception. In any

event, as long as the main objections to the conception remain unan-

swered, the opposing conception supplied by perfectionism should seem

more promising. Being a conception that has no trouble distinguishing

human from bovine well-being or excluding a priori the possibility of the

life of total passivity�s being the best human life, it avoids the embarrass-

ments of the hedonistic conception. For this reason at least, it appears to

yield the sounder form of eudaimonism.


6. Plato�s ethics

Let us continue to take Plato�s ethics as the touchstone for this form of

eudaimonism. Recall, from our earlier account, that Plato developed his

ethics from three related theses. First, the highest good for a human being is

achieved through the development of awell-ordered personality. Second, in

such a personality, the constituent parts, reason, spirit, and appetite, work

together harmoniously. And third, this harmony of mind or soul results

when reason, with the help of the emotional energy of spirit, regulates and

constrains appetite. Plato placed reason at the helm of the well-ordered

personality because, as he put it, good governance requires knowledge of

what things benefit and harm the governed, and reason is the seat of all

knowledge. This knowledge, then, that good governance requires is the

same as wisdom. Accordingly, wisdom is the virtue a person possesses

when reason excels in its capacity as the ruling part of the soul. Further, it

consists both in the knowledge of what things are worthwhile in them-

selves and in the knowledge of how best to achieve them given one�s

circumstances. This knowledge is the most important kind a person can

have, though of course it is not the only kind. Other kinds fall into the

category of technical expertise, the special knowledge that artisans and

technicians, such as builders and navigators, have. Such knowledge equips

its possessors with the command of their specific occupations that is neces-

sary to be successful in them. In our examples these are the activities of

building houses and moving ships across the sea. Wisdom, by contrast,

equips its possessors with the self-command necessary to be successful in

the much more general activities of living a life and leading a community.

Wisdom is the first virtue in Plato�s ethics. On his theory, because one�s

highest good is the perfection that one achieves through the development

of a well-ordered personality, and because such perfection results from

reason�s excelling in its governance of one�s soul, it results from wisdom.

To be sure, these points alone do not distinguish Plato�s conception of

wisdom from conceptions that hedonistic forms of eudaimonism could

advance. Epicurus, too, could hold that wisdom was the first virtue, for he,

too, took wisdom to be general knowledge of what things benefit and harm

one and thought that one achieved the highest good by acquiring such

knowledge and putting it into practice. What distinguishes Plato�s concep-

tion from Epicurus� is the special object of the knowledge that Plato

Eudaimonism 71

identified with wisdom. Where Epicurus took the object of the knowledge

he identified with wisdom to be pleasure and its sources, Plato took it to be

the idea of perfection itself. Consequently, while a wise person, on a theory

like Epicurus�, regards his wisdom and the other virtues that follow from it

as useful to the achievement of the highest good, he does not regard them as

constitutive of that good. By contrast, on Plato�s theory, since wisdom and

the other virtues are themselves perfections, they exemplify the very idea

that is the object of wisdom, and hence are regarded by the wise person as

constitutive of the highest good. That good, in other words, on Plato�s

theory, not only is achieved through wisdom but also contains wisdom

and the other virtues as integral parts. Such is the consequence of Plato�s

taking perfection rather than pleasure as the highest good.

The wise person, according to Plato, both comprehends and finds com-

pelling the idea of perfection. He is like amaster craftsman or accomplished

artist whose knowledge of his craft or art and corresponding grasp of

perfection in its exercise guide him toward the production of as perfect an

artifact, composition, or performance as he can create. Wisdom, therefore,

for Plato, is knowledge of perfection itself and how it applies to the conduct

of a human life and also to the life of a city. In the latter case, a city displays

wisdom when its rulers have such knowledge and apply it in governing the

community of people of which their city consists. Such excellence in state-

craft thus consists in the rule of philosophers and its acceptance by their

auxiliaries and the working men and women who make up the majority of

the city�s population. And Plato, then, following his analogy between the

city and the soul, identifies wisdom in the individual with the rule of reason

and its acceptance by spirit and appetite when reason has attained the

requisite knowledge. Wisdom is thus exemplified in a well-ordered person-

ality by the excellence of reason, having attained knowledge of perfection,

in governing how one conducts one�s life.

By the same token, Plato identifies the other virtues exemplified in a

well-ordered personality by their place in the operations of its three con-

stituent parts. He identifies courage, for instance, with the workings of

spirit under the guidance of reason. Courage, on Plato�s view, consists in

spirit�s working to supply the confidence and strength of character one

needs to follow reason�s direction when one is faced with threats from

which a person of lesser mettle would shrink. This view nicely matches

the traditional view of courage as the power to stand up to threats when


pursuing goods of greater value than that of safety from those threats,

which the easily frightened are prone to seek. Spirit, as Plato understands

it, is like the flight–fight mechanism that ethologists attribute to animals in

explaining their responses to predators and rivals: an inherited mechanism

that when properly trained and brought under the command of reason

enables its possessor to stand fast in defense of important goods when he

is threatened with their loss and to persevere in their pursuit despite the

dangers that the pursuit involves. Similarly, the virtue of moderation, Plato

points out, is exemplified in the well-ordered personality by appetite�s

subordination to reason. Traditionally, moderation is the power to resist

strong desires, like those for rich food, intoxicating drink, and carnal

pleasure, whose indulgence would bring one more harm than good, and

again Plato�s identification of this virtue with appetite�s submission to

reason nicely matches this traditional view.

Of course, the virtue that is of special interest to Plato is justice. He finds

this virtue, too, in the well-ordered personality. But the question is where,

since unlike the other three virtues, wisdom, courage and moderation,

justice is not particularly associated with the workings of any one of the

soul�s constituent parts. Recognizing this, Plato again looks to his analogy

between the constitution of a well-governed city and that of a well-ordered

personality. The strategy makes good sense if one accepts the analogy. For

justice must exist in the well-governed city. No well-governed city, after all,

could also be unjust. Further, what makes a city well-governed, Plato

observes, is the fidelity of each of its constituent parts to fulfilling its proper

role. So justice, according to Plato, as an excellence of a city and, and indeed

of any political community, consists in the community�s parts each playing

its proper role in harmony with the others. Correspondingly, Plato then

argues, justice as a personal virtue consists in the constituent parts of the

soul each playing its proper role in harmony with the others. Justice, there-

fore, is exemplified in a well-ordered personality, just as it is exemplified in

a well-governed city. And once one locates its source in the latter, one can

see it in the former as well.

We have now arrived at Plato�s solution to the Republic�s core problem. It

offers not only a very congenial picture of the best life for human beings

but, even more important, reassurance that such a life is possible only for

those who live justly. Crooks and bullies, Plato tells us, necessarily live

inferior lives, for having turned to injustice to achieve their ends, they

Eudaimonism 73

have abandoned the possibility of achieving the highest good and settled

instead for somethingmuch cheaper. This is unquestionably an elegant and

inspiring solution. It shows not only that justice, being integral to a well-

ordered personality, is good in itself and not merely good as a means to

something else, but also that justice, because it is necessary to achieving the

highest good, is superior in value towhatever good one can achieve through

injustice. Thus it speaks directly to the challenge that Glaucon and

Adeimantus put to Socrates.

Nonetheless, the solution is open to a powerful objection. The analogy

between a well-governed city and a well-ordered soul on which it depends

does not get Plato all the way to the conclusions he needs to answer

successfully Glaucon and Adeimantus� challenge. It fails because, though

it supports attributing justice to the well-ordered soul, the attribution it

supports is only of justice in the relations among the soul�s constituent

parts and not of justice in the way an individual whose soul is well-ordered

acts toward others. One can see this, once one recognizes the limits onwhat

the analogy implies. Specifically, the justice it supports attributing to awell-

ordered soul must be like the justice that is necessarily found in a well-

governed city on account of its being well-governed, and the latter consists

in justice among the members of the city in their conduct toward each

other. In particular, it is realized in government by those who are equipped

to rule, helped in their efforts by auxiliary forces, and obeyed by the

members of the city�s economically productive class, who, knowing the

wisdom of their rulers, willingly accept their authority. Correspondingly,

the justice that one may attribute to a well-ordered soul on the basis of this

analogy consists in justice among the soul�s constituent parts in their

relations to each other. It is realized, that is, in the rule of reason, when

reason through education is equipped to rule and when its rule is backed by

the workings of spirit and unchallenged by appetite. The existence of such

justice in the soul, however, does not imply that someone whose soul is just

in this way will act justly towards others. Nothing, for instance, in this

characterization of justice as a personal virtue implies that one whose

soul is well-ordered is certain upon finding a purse containing a huge wad

of cash to try to return it, and the cash, to its owner.

The gap in Plato�s argument is due to Plato�s having failed to take note of

two different areas of political life in which a city can exhibit justice. First, it

can exhibit justice in its domestic or internal affairs. It exhibits such justice


when the relations among itsmembers are wholly just. A well-governed city

necessarily exhibits such justice, for it cannot be true of a city that it is well-

governed if the relations among its members are unjust. Second, a city can

exhibit justice in its foreign or external affairs. It exhibits such justice when

its conduct toward other cities is just, when, for instance, it fully complies

with the trade agreements it has entered into with them and does not seek

unfair advantage by cheating on those agreements. Such justice, though it

may, as a rule, be observed by well-governed cities, is not necessarily

observed by them. That a city is well-governed, in other words, is no guar-

antee of its being just in its foreign or external affairs. For its being well-

governed means that eachmember keeps to his or her proper role and does

not cause any internal strife by trespassing on other members� roles, and a

city that behaves unjustly toward other cities may nevertheless have

achieved such civic harmony. Plato�s error, then, consists in his conflating

civic justice in domestic affairs with civic justice in foreign affairs, for his

analogy between a soul and a city allows him to attribute the analog of the

former to a well-ordered soul but not the analog of the latter. Yet it is the

analog of the latter that awell-ordered soulmust have for possession of such

a soul to imply that one is just in one�s conduct towards others. It is the

analog of the latter, that is, that a well-ordered soul must have if Plato�s

account of it as the highest human good is to yield a successful answer to

Glaucon and Adeimantus� challenge.

Though Plato�s solution to the Republic�s core problem falls short of

answering Glaucon and Adeimantus, it contains an important insight that

suggests other, possiblymore promising solutions. The insight is implicit in

Plato�s account of the motives behind unjust action. On a plausible recon-

struction of that account, Plato proceeds on the unstated assumption that

injustice is never an ultimate end of a person�s actions. Consequently, when

injustice is done, it is done either as a means to some other end, as embez-

zlement is a means to enrichment, or as a side effect of an action whose

intended effect is different, as depriving others of their fair share is a side

effect of grabbing things one wants before anyone else can take them. In

either case, themotive behind the act, Plato thought, is an unmet need such

as an appetite in need of gratification or a fear in need of removal. People, in

other words, are moved to act unjustly by needy states, states that demand

satisfaction or relief. When such states are the dominant motives in a

person�s life, then that person is ruled by his appetites and fears, and his

Eudaimonism 75

life consists in striving for gratification of the former and escape from the

latter. The striving is ceaseless, short of death, though itmay occasionally be

punctuated by temporary states of satiation and repose. Thus the ends that

give shape and direction to this life of ceaseless striving are nevermore than

temporarily achieved. In such a life, situations recommending injustice as

the expedient course of action occur repeatedly, yet whether one chooses

expedience and so acts unjustly or resists and so acts justly makes little

difference in the long run. The life is unfulfilling in either case. Such is the

fate of the person inwhomappetite and its correlative fears dictate his ends.

Plato�s way out of this quandary is to find in our personalities a source of

ends that, unlike those whose source is appetite and its correlative fears, are

permanently achievable. These are the ends of perfection. Their source is

reason. With the proper education of reason, Plato held, a person can come

to understand perfection in itself, to know its many forms, beauty, truth,

goodness, and justice among others, and to commit himself deeply to real-

izing them in his life. Such a person would then live a life in which the ends

of perfection dominated thosewhose sourcewas appetite and its correlative

fears. The former ends and not the latter, that is, would shape and give

direction to his life. Accordingly, such a person would be ruled by reason

rather than appetite, and his life would consist in contemplative and crea-

tive activities through which he could realize perfection. Such a life, Plato

believed, would be complete and fulfilling. In particular, it would lack the

bottomless privations characteristic of life lived under the rule of appetite.

And being free of those privations, people who lived such lives would never

see their situations as recommending injustice as the expedient course of

action. They would always act justly as a matter of course. The reason why,

Plato thought, is that the realization of perfection entails the development

of a well-ordered personality and therefore, by analogy to the well-governed

city, the instantiation of justice as an integral aspect of one�s personality. Of

course, as we saw, this argument of Plato�s fails. The analogy does not imply

that people who have developed a well-ordered personality will act justly

towards others. But the general ideas leading to it, that people whose lives

were complete and fulfilling would not be liable to the motives from which

unjust actions towards others spring and that such lives are possible only if

shaped and directed by ends whose achievement can be permanent and so

by ends that are distinct from those of appetite and its correlative fears, are

nonetheless promising. To see whether they can yield a sound solution to


the Republic�s core problem, however, wemust look to a different account of

human personality from Plato�s.

7. Rationalism v. naturalism

Plato�s ethics initiated the great tradition of rationalism in the theory of

moral psychology. Its central themes are the fundamental opposition

between reason and appetite (likewise reason and passion) and the suprem-

acy of reason in any well-functioning and upright personality. In this tradi-

tion humanmotives are divided into two kinds: themotives that the powers

of rational thought and action produce and the motives that come from

animal appetite and brute emotion. Adherents to the tradition then charac-

terize human life as a site of struggle between motives of the former kind

andmotives of the latter and as livedwell onlywhen the former prevail over

the latter. The nearly universal struggle to curb an appetite for sweets and

the now familiar struggle to suppress a desire to get back at some incon-

siderate driver would be typical examples. Each arises from an immediate,

seemingly instinctive attraction or impulse, to which, knowing of the

harmful effects of eating too many sweets or the dangers of petty tit-for-

tat behavior, one responds with resistance, and this resistance, on the

rationalist tradition, represents the force of reason. Accordingly, such strug-

gles are seen as revealing a basic division in human personality between its

rational and irrational parts and indicating the need for the former to

control and rule the latter. The ideal rationalist resolution thus consists in

reason�s mastering appetite and passion and so in the subordination of the

latter to the rule of the former. In all of these ideas and themes rationalism

has followed Plato�s lead.

Its chief opponent in the theory of moral psychology has been natural-

ism. This is because in the rationalist tradition, again beginning with Plato,

reason is seen as operating independently of the forces of nature. In the

naturalist tradition, by contrast, the operations of reason are understood to

be natural phenomena explicable, like other natural phenomena, by the

forces and conditions of nature. The foremost point of disagreement

between the two traditions, then, is over what I will call the doctrine of

the independence of reason, which rationalists affirm and naturalists deny.

Other points of disagreement follow from this one, at least when ration-

alists, as they have commonly done, interpret this doctrine as having

Eudaimonism 77

certain metaphysical implications. Thus rationalists and naturalists have

traditionally disagreed over the nature of death and the possibility of the

immortality of the soul, for rationalists have commonly interpreted the

doctrine of the independence of reason as implying that the human soul,

in virtue of its being the seat of reason, can exist outside of the natural world

and therefore independently of the human body it inhabits. And when so

interpreted, the doctrine supports a conception of death as consisting in the

separation of the soul from the body and the consequent belief in the

possibility of the soul�s being immune to the destructive forces of nature.

Naturalists, by contrast, having denied the doctrine of the independence of

reason, find nothing in the mental activities and processes of human life to

imply that any part of the human soul exists outside of nature or that it

enjoys an immunity from nature�s destructive forces. According to natural-

ism, these activities and processes are themselves activities and processes of

life. They are no different in this respect from respiratory activities and

processes. Hence, they cease when life itself ceases, if not before, and death

consequently, as naturalists conceive of it, consists in their permanent

cessation.

These disputes show the metaphysical depths of the traditional opposi-

tion between rationalism and naturalism. At the surface, however, the

disputes between the two traditions concern matters of psychology alone.

Above all, they concern the possibility of reason�s being a source of motives

that can conflict with themotives that come from animal appetite and brute

emotion. Naturalists deny that it can. They deny, that is, that the operations

of reason can ever generate motives. The powers of reason, they maintain,

do not include the power to initiate action. Thus they deny that when,

knowing of the harmful effects of eating too many sweets, say, we resist

the attraction of after-dinner chocolates, our resistance represents the force

of reason. What it represents instead, they argue, is the fear of pain and

disability that, owing to our foreseeing the pain and disability caused by

eating too many sweets, moves us to resist the temptation. And they trace

this fear to its origins in the aversion to pain and the instinct for self-

preservation that human beings, like other sentient animals, have inherited

as part of their motivational constitution. In this way they exclude reason

from being a source of motives. On naturalist theories of moral psychology,

reason contributes to the formation of motives, not by originating them,

but by providing the intelligence necessary to guide actions motivated by


fear and aversion, desire and inclination, away from objects of the former

and toward objects of the latter. It serves, in other words, as a pilot of action

only, and not as an engine. Hence, in direct opposition to the rationalist idea

of the conflict between reason and appetite or reason and passion, natural-

ists maintain that all conflicts of motives are conflicts among appetites and

passions. Some of these motives, being informed by foresight and knowl-

edge, are more conducive to one�s well-being and to the promotion of the

highest good than others, and what rationalists take to be conflicts between

reason and appetite or reason and passion are therefore, on naturalists�

theories, conflicts between these enlightened motives and motives that are

uninformed by foresight and knowledge.

We might, then, look to this naturalist tradition for the nonplatonic

account of human personality that we seek in considering whether Plato�s

ideas about themotives of injustice can yield a sound, eudaimonist solution

to the Republic�s core problem. To see whether they can, let us first configure

Plato�s ideas to fit these naturalist accounts. In particular, we need to

replace Plato�s account of the human personality as divided into rational

and irrational parts that can produce conflictingmotives with an account of

human personality on which our rational powers do not include the power

to produce motives. That is, instead of working with conceptions of reason

and appetite as being separate sources of motivation that can oppose each

other, we must work with conceptions of them as being separate capacities

with different functions, cognitive in the case of reason, motivational in the

case of appetite, that work together more or less coordinately in the pro-

duction of human action. Accordingly, appetite supplies motivational

energy that powers human action, and reason directs that energy and so

guides the movements it powers. Such direction and guidance come about

through the greater understanding of what a good life consists in and how

best to achieve it that reason provides. What makes for a well-ordered

personality, then, on this naturalist account, is the harmonious coordina-

tion of reason�s cognitive functions with appetite�s motivational functions.

When one�s judgments about what activities and experiences one ought to

pursue perfectly fit the ends one is motivated to pursue, then reason and

appetite are working together harmoniously. Specifically, when one�s judg-

ments of the relative importance of some activity or experience to advanc-

ing one�s well-being matches the strength of one�s motivation to engage in

that activity or pursue that experience, then one�s personality is well-

Eudaimonism 79

ordered. We can thus define, on the naturalist account, an ideal of a

well-ordered personality that corresponds to the one to which Plato�s argu-

ment appeals.

At the same time, wemust keep in mind how it departs from the ideal to

which Plato�s argument appeals. Plato, recall, makes the development of a

well-ordered personality dependent on one�s attaining knowledge of

perfection. He identifies such knowledge with wisdom and maintains that

to attain it, reason must undergo a rigorous education in the disciplines

whose mastery is necessary to understanding perfection in all of its forms.

Once this education is successful, on Plato�s account, the rational part of the

soul will have achieved supremacy over the soul�s other parts and will, with

the aid of spirit, rule appetite. The personality will be well ordered. On the

naturalist account, by contrast, development of a well-ordered personality

depends not only on the attainment ofwisdombut also on the acquisition of

the self-command necessary to act in accordance with the deliverances of

wisdom. Hence, in addition to the education of reason, by which one

acquires wisdom, appetite, too, on this account, must be conditioned so

that the ends toward which it moves one and the strength of one�s desire to

realize those ends match what wisdom determines are the activities worth

doing and the experiences worth pursuing. To achieve harmonious coordi-

nation of reason�s cognitive functions with appetite�s motivational func-

tions requires, in otherwords, training the latter so as to enable one to enact

the wisdom that results from the education of the former. Only then will

one acquire, in addition to wisdom, the excellences of courage,moderation,

and justice, among others. A well-ordered personality, on the naturalist

account, unlike Plato�s, does not, then, result from the successful education

of reason alone.

By the same token, one cannot, on the naturalist account, unlike Plato�s,

assume that the proper education of reason guarantees the acquisition of

the ends whose pursuit makes for a complete and fulfilling life. Plato can

make this assumption because reason, on his account, is the source of such

ends by virtue of being the source of themotives whose dominance over the

motives of appetite is the essence of a well-ordered personality. But on the

naturalist account, because reason is not the source of such motives or

indeed any motives, one cannot assume that it is the source of such ends.

What the ends are whose pursuit makes for a complete and fulfilling life is

itself a question for naturalists. They cannot, that is, attribute such ends to


reason without argument. Further, for naturalists who propound perfec-

tionist theories of human well-being, the question of what the ends are

whose pursuit makes for a complete and fulfilling life is a question of what

activities in life are worth doing for their own sake. We must therefore find

an answer to this question before we can determine whether the accounts

of human personality that such naturalists put forward offer a solution to

the Republic�s core problem.

8. Aristotle�s naturalism

This naturalist program goes back to Aristotle (384–322 BC). After Plato,

Aristotle is the most important exponent of perfectionism in the theory of

well-being. He is also, after Plato, the most important exponent of the

specific form of eudaimonism that Plato originated. Like Plato, Aristotle

identifies human well-being (eudaimonia) with the ideal of living a com-

plete and fulfilling life. It is complete, Aristotle says, in the sense that one

could not make it a better life through the addition of some further good,

and it is fulfilling in the sense that one aims to live such a life for its own

sake and not for the sake of some other good. In short, such a life lacks

nothing that could improve it and is wholly satisfying in itself. It thus fits

Plato�s ideal of a life lived in pursuit of ends that are permanently achiev-

able. But unlike Plato, Aristotle does not identify these ends with perfection

itself. He takes ethics to be a study of the good for human beings without

supposing that it proceeds from a study of goodness or perfection generally.

Thus, while he agrees with Plato about the completeness and self-

sufficiency of the best human life, he differs from Plato in thinking that

knowledge of what such a life consists in requires attention to the nature of

our humanity and to the sort of life that is distinctive of our species and not,

as Plato thought, to some abstract idea of perfection by which everything

that is good in itself is measured. And while Aristotle also shares Plato�s

general view that the best human life is the life of rational thought and

activity, he bases this view on decidedly different considerations from those

on which Plato based it.

Specifically, Aristotle bases it on considerations of biology. That is, he

develops the view by considering human beings as living things. As such,

he holds, humans have a distinctive form of life. Accordingly, determining

what makes for a good human life requires first defining its distinctive

Eudaimonism 81

form, just as one would first define, say, what the distinctive form of

literature a novel is before determining what makes for a good novel or

the distinctive form of meal a buffet luncheon is before determining what

makes for a good buffet luncheon. And the way Aristotle defines the form

of life distinctive of humankind, or indeed of any animal species, is to

identify those powers that are special to that species and then to make

the exercise of those powers definitive of that form of life. From such a

definition it follows that the best life for the members of the species entails

full development of those powers and excelling in activities that display

their exercise.

Consider, as an illustration of this approach, how one would determine

whatmakes for the best life for eagles. To answer this question requires first

defining the form of life distinctive of eagles, which means identifying the

eagle�s special powers. These include, not only the power of flight, but also,

in particular, the power to glide for great distances high above the ground.

Further, they include remarkable visual powers, specifically, the power to

see faraway objects clearly and distinctly. Accordingly, the form of life

distinctive of eagles includes such activities as dwelling on high cliffs or

other eyries, soaring over large tracts of land, and homing in on prey with

great accuracy. The best life for an eagle therefore entails living in this way

and excelling at these activities. Eagles that live in other ways necessarily

have worse lives, for they have lives in which their distinctive powers are

either dormant or exercised weakly and defectively. Eagles that live in

captivity, for instance, whose wings are clipped and whose nests are near

to the ground, however well-fed, disease-free and lovingly cared for they

may be, live markedly worse lives than those that thrive in their natural

habitat. For such eagles never exercise the powers that are the special gifts

of their species. Indeed, their existence, when regarded in this light, seems

sad. Though they are eagles, their lives seem stunted by comparison with

the form of life that we readily identify as an eagle�s.

For Aristotle the same sort of consideration applies in defining the form

of life distinctive of humankind and determining, in consequence, what is

the best life for human beings. Unsurprisingly, Aristotle singles out reason

as the power special to human beingswhose exercise defines the formof life

distinctive of our species. So the best life for human beings, Aristotle main-

tains, is a life in which one�s rational powers are fully developed and in

which one excels in activities that display their exercise. Men and women


who never develop their powers of reason or whose lives lack opportunities

for exercising them, people who cannot without assistance deal with the

complex circumstances that are common to life in all but perhaps the most

primitive human societies or who work long hours in mindless, repetitive

tasks, do not have as good lives as those whose rational powers are fully

developed and whose occupations and avocations afford plenty of oppor-

tunity for their exercise. The former, on Aristotle�s view, have a sadder

existence in comparison to the latter, just as eagles who live in captivity

have lives that pale in comparison to those of their conspecifics who live in

the wild. This is not to say, of course, that their lives are less pleasurable,

though itwould be astonishing if theyweren�t. For Aristotle�s theory ofwell-

being, remember, is that of perfectionism and not hedonism. It is to say,

then, that their lives fall significantly short of the perfectionist ideal of

human well-being as compared to the lives of the latter.

To be sure, to characterize the best life for human beings as a life in which

one�s rational powers are fully developed and inwhich one excels in activities

that display their exercise is not yet to identify ultimate ends in life whose

pursuit makes for a complete and fulfilling life. A great many different kinds

of lifefit this characterization, for the range of human activities excellence in

which requires the exercise of one�s rational powers is indefinitely large.

What is more, plenty of these activities have ends whose pursuit would not

make for a complete and fulfilling life if theywere one�s ultimate ends in life.

A life spent in pursuit of fame, the life of the publicity seeker, for instance,

though it would require the exercise of rational powers, would not be com-

plete and fulfilling if fame were its dominant end, for it is in the very nature

of fame to die out over time in the absence of renewed efforts at gaining

publicity. Hence, this characterization of the best life for human beings is not

sufficient for determining the ends whose pursuit makes for a complete and

fulfilling life. Aristotle, therefore, since he meant his view of the best life for

human beings to be one of a life that is complete and fulfilling, must have

thought there was more in the considerations of biology on which he based

this view than we have so far allowed.

In fact, Aristotle did think there was more. His view, which he expressed

in an important passage in his Nicomachean Ethics, is that the very point of

human life is to exercise reason.12 Rational activity for its own sake,

12 Aristotle, Nicomachean Ethics, bk. I, ch. 7, 1097b22–1098a20.

Eudaimonism 83

Aristotle held, is therefore the end whose pursuit makes for the best life for

human beingswhen one excels in that activity. This conclusion, he thought,

follows from analogy. Just as the point of being a flute player is to play the

flute and the point of being a sculptor is tomake sculptures, so too there is a

point of being human. Or as Aristotle put it, in drawing a second analogy,

just as eyes have a natural function and hands have a natural function, so

too in nature the whole of man has a function. And to understand what that

function is, one must determine the powers and capabilities that distin-

guish humankind from other animal species. Aristotle, as we noted before,

took these to be the powers of reason. Hence, for him the same consider-

ations that lead to defining the form of life distinctive of humankind also

determinewhat humankind�s function is. On his view, then, the best life for

human beings is a life in which one fulfills that function, and such a life

would be one whose point and final end is rational activity. Such a life, if it

displayed excellence in the activities to which it was devoted, would neces-

sarily be complete and fulfilling, for a life in which one fulfilled one�s

function as a human being and achieved excellence in doing so would be

one that was wholly satisfying in itself and lacked nothing that could make

it better.

How arewe to understand this argument of Aristotle�s? Itsmeaning is not

immediately clear, and the obscurity is undoubtedly due to the distance

between the biological notions of the ancient Greeks and the notions of

modern biology with which we are familiar. Still, the analogies on which

Aristotle bases his thesis that human beings have a distinctive function in

virtue of being human offer some clue to understanding his argument.

Consider, then, his appeal to flute players and sculptors. To be either, one

must engage in the activity that is distinctive of their art.When flute players

or sculptors engage in this activity they function as flute players or sculp-

tors, as the case may be, and this functioning defines their function. So too,

Aristotle must be arguing, when a man engages in the activity that is

distinctive of his species, he functions as a human being, and this function-

ing defines his function. To function as a human being, according to

Aristotle, is to engage in rational activity, for reason is the power that

distinguishes human beings from other animals. Looking and listening,

smelling and tasting, by contrast, are activities that both humans and

other animals engage in. Consequently, one functions as a sentient animal,

and not particularly as a human being, when one looks at things, listens to


things, smells them, or tastes them. The sensory powers one exercises in

these activities are not peculiar to human beings. Likewise one functions

merely as a living thing, and not specially as a human being or even as a

sentient animal, when one drinks, eats, and sleeps, for such nutritive and

metabolic activities are common to both plants and animals. And because

Aristotle expressly draws these contrasts, one can assume that he means to

distinguish between activity whose very point is the exercise of reason and

activities whose point is the exercise of these other powers. Hence, what he

means to show by this argument is not that only some of the ends whose

pursuit requires the exercise of reason are ends whose pursuit canmake for

a complete and fulfilling life. Rather, he means to show that any set of ends

whose pursuit requires the exercise of reason are ends whose pursuit can

make for a complete and fulfilling life if the point of pursuing them is

ultimately to develop and exercise one�s reason, which is to say, to live a

life in which one functions as a human being, and to achieve excellence in

so functioning.

Let us call this the broad interpretation of Aristotle�s function argument.

It is broad because the argument, on this interpretation, is not meant to

narrow the range of ends around whose pursuit a life that has the form

distinctive of humankind must be organized to be the best life for a human

being. Rather it is meant to establish that a certain view and attitude toward

those pursuits is necessary to living a complete and fulfilling life. A person�s

life becomes organized around the pursuit of certain ends long before he

comes to reflect on the meaningfulness of pursuing them. Success in life

consists in his achieving those ends, so achieving them, he may naturally

think, on first reflection, is what makes their pursuit meaningful. Later,

though, if the pursuit of these ends calls for the development and exercise

of his rational powers, he may come to see that his having such ends has

created for him a life whose form is distinctively human and so to regard the

activity that qualifies it as distinctively human as what makes their pursuit

meaningful. He will have come, in other words, from seeing and valuing

these pursuits as means to success in life to seeing them as consisting in

activity whose point is the exercise of reason essential to it and whose value

therefore lies in his excelling in such rational activity. He will thus have

come to have the view and attitude toward the pursuits around which his

life is organized that, on the broad interpretation of Aristotle�s function

argument, is necessary for his life to be complete and fulfilling, for only

Eudaimonism 85

then will the point of these pursuits have become for him the development

and exercise of his rational powers so as to excel in his functioning as a

human being. Only then will he be living a life that meets Aristotle�s

definition of the best life for a human being: the life consisting in activity

of human thought and action in accordance with virtue. And with this

definition, Aristotle comes to the final conclusion of his function argument.

9. A problem in Aristotle�s program

At this juncture, however, Aristotle�s naturalist program runs into trouble.

The trouble occurs because Aristotle takes human virtue to consist in the

capabilities and dispositions whose exercise corresponds to excellence in

human functioning, and he identifies these with all the standard virtues,

wisdom, courage, moderation, justice, honesty, generosity, and the like. In

short, he believes that excellence in human functioning consists in activ-

ities that call for the exercise of these standard virtues. Yet his function

argument, on the broad interpretation of it, is too broad to support this

belief.

The reason is that the argument, on this broad interpretation, does not

exclude any life that has the form distinctive of humankind from being the

best life for human beings on account of the ends aroundwhose pursuit it is

organized. Even a life spent in pursuit of fame is consistent with being the

best life for a human being, for it is conceivable, however implausible as a

matter of human psychology, that a person who spent his life in pursuit of

fame could see his pursuits as affording him opportunity to exercise his

rational powers and therefore take the point of the activity that his pursuit

of fame consisted in to be the development and exercise of those powers so

as to excel in his functioning as a humanbeing. So too, a life spent swindling

others, the life of a con artist, as it were, is, as far as Aristotle�s function

argument goes, on its broad interpretation, consistent with the best life for

human beings, for it, too, is a life success inwhich requires the development

and exercise of one�s rational powers. And it is therefore conceivable,

however implausible as a matter of human psychology, that a person who

spent his life swindling others could regard the pursuits around which he

had organized his life as affording him opportunity to exercise his rational

powers and therefore take the point of the activity that the pursuit of a good

con consisted in to be the development and exercise of those powers.


Clearly, though, such a life is inconsistent with a person�s having the virtues

of justice and honesty. Hence, the function argument, on its broad inter-

pretation, does not support Aristotle�s belief that functioning well as a

human being consists in activities that call for the exercise of all the stand-

ard virtues, including, in particular, justice and honesty.

To support Aristotle�s belief, then, the function argument has to be given

a different interpretation. Specifically, one has to interpret it as being

meant to narrow the range of ends around whose pursuit a life that has

the form distinctive of humankind must be organized to be the best life for

human beings. And to interpret the argument in this narrower way means

taking Aristotle as holding that some forms of rational activity are, by virtue

of their forms, better suited to realizing excellence in human functioning

than others. On one such narrower interpretation, for instance, Aristotle

holds that the pursuit of truth and knowledge elicits the development and

exercise of one�s rational powers to a higher degree of perfection than the

pursuit of fame, wealth, public service, triumph in battle, or artistry in any

art or craft. Accordingly, the best life for human beings would be a life

which was organized around the pursuit of truth and knowledge and in

which one excelled in the activities essential to discovering truth and

advancing knowledge, which is to say, science and philosophy, research

and scholarship.

Plainly, such an interpretation of Aristotle�s function argument, and

indeed any interpretation of it as being meant to narrow the range of ends

around whose pursuit a life that has the form distinctive of humankind

must be organized to be the best life for human beings, depends on attrib-

uting to Aristotle a conception of our rational powers on the basis of which

one can distinguish some forms of rational activity from others as truer or

more exemplary exercises of those powers. For otherwise one could not

interpret him as holding that some forms of rational activity were, just by

virtue of their forms, better suited than others to realizing excellence in

human functioning. Yet it is difficult to see how one could justify drawing

such a distinction among forms of rational activity. And if one cannot, then

the function argument must fail to exclude from being the best life for

human beings a life that is organized around ends whose pursuit is incon-

sistent with having all the standard virtues.

The problem with justifying such a distinction is this. Human intelli-

gence comprises a number of different capacities for productive thinking.

Eudaimonism 87

These include capacities for conceptual and logical thought, for invention,

for creative synthesis of sensory experiences, for empathic understanding,

andmore. A conception of our rational powers that can serve as the basis for

distinguishing some forms of rational activity from others as truer or more

exemplary exercises of reasonmust be one that either limits reason to some

one or few of these capacities or, though it comprehends all of them,

sharply distinguishes each from the others. Yet clearly, if the latter is the

conception of our rational powers that is attributed to Aristotle on the

narrower interpretation of his function argument, then the argument

must also include a premiss to the effect that, though the exercise of any

of these capacities is an exercise of reason, the exercise of some is a truer or

more exemplary exercise of reason than the exercise of others. And such a

premiss would be very hard to justify. To elevate any of these capacities

above the others as a capacity whose exercise represents a purer or more

perfect formof reason, when one takes the exercise of each to be an exercise

of reason, would seem merely to reflect some arbitrary preference for one

form of rational activity over others.

Suppose, then, that the former is the conception of our rational powers

that is attributed to Aristotle on the narrower interpretation of his function

argument. It follows that our rational powers, on this second version of the

interpretation, include some but not all of the capacities for productive

thinking that human intelligence comprises. Consequently, the conception

can serve as the basis for distinguishing some forms of rational activity from

others as being truer ormore exemplary exercises of reason, for one can use

it to distinguish the elements of the activity that are due to exercises of

reason from the elements that are due to exercises of other capacities that

human intelligence comprises. Accordingly, one form of rational activity is

a truer or more exemplary exercise of reason than another the greater the

contribution of the powers of reason to that activity relative to the contri-

butions of these other capacities. But now what had appeared, on the first

version of the narrower interpretation, to be an arbitrary preference for one

form of rational activity over others reappears, on this second version, as an

arbitrary preference for one form of human functioning over others. After

all, if some of the capacities for productive thinking that human intelli-

gence comprises are distinct from our rational powers, then Aristotle has to

justify excluding their exercise from humankind�s distinctive function, and

it is hard to see how he could. Since there appears to be no basis for taking


only our rational powers as what distinguishes humankind from other

animal species, when human intelligence is understood to comprise other

capacities for productive thinking, there appears to be no basis for taking

the exercise of reason to be humankind�s function to the exclusion of these

other capacities.

The upshot is that Aristotle�s function argument, on either version of the

narrower interpretation, does not support taking the exercise of any of the

capacities for productive thinking that human intelligence comprises as

more excellent or perfect kinds of human functioning than the exercise of

any other. Hence, it does not support taking one life as better for human

beings than another because pursuit of the ends aroundwhich the former is

organized is more conducive to the exercise of some one of these capacities

than pursuit of the ends around which the latter is organized, given that

pursuit of the ends around which the latter is organized is more conducive

to the exercise of a different capacity among those that human intelligence

comprises. The argument does not, for instance, support taking a life

devoted to the pursuit of truth and knowledge as a better life than a life

devoted to music, painting, cooking, or some other fine or practical art.

While the former life is presumably organized around endswhose pursuit is

more conducive to the exercise of the capacity for conceptual and logical

thinking than pursuit of the ends around which the latter is organized,

nothing in the argument supports taking the exercise of that capacity as a

more excellent or perfect kind of human functioning than the exercise of

the capacity for synthesis of sensory experience, the capacity to whose

exercise pursuit of the ends around which the latter life is organized is

more conducive. And similarly for a life organized around ends whose

pursuit is most conducive to exercising the capacity for invention: the life

devoted to solving practical problems, making new tools, or organizing

cooperative enterprises. Indeed, the same holds even for a life that is

organized around ends whose pursuit is most conducive to exercising

the capacity for seeing absurdity and incongruity, the life devoted to

comedy, parody, mime, or other forms of humorous entertainment and

commentary.

In sum, Aristotle�s function argument, if interpreted as meant to limit

the range of ends aroundwhose pursuit a life that has the formdistinctive of

humankind must be organized to be the best life for human beings, is too

weak to establish such limits. And failing this, the argument cannot reach

Eudaimonism 89

the conclusion that Aristotle intended. His function argument notwith-

standing, a man could live a complete and wholly satisfactory life, a life in

which he fulfilled humankind�s function and achieved excellence in doing

so, yet the thought and action to which he devoted his life need not be in

accordance with all the standard virtues, justice and honesty in particular.

10. Prospects for contemporary eudaimonism

Like Plato, Aristotle too believed that people whose lives were complete and

fulfilling would not be liable to the motives from which unjust and dishon-

est actions spring. Whether or not he agreed with Plato in seeing these

motives as needy states, he did regard them as states of excessive appetite

and passion that dispose one to seek more than one�s due. As such, they are

dysfunctional states. To act under their influence is to act badly, which is to

say, defectively as a human being. Consequently, to live a life in which one

fulfills one�s function as a human being and achieves excellence in doing so

is to live a life free of these dispositions. Justice, Aristotle thought, is there-

fore among the virtues someone who lived a complete and fulfilling life

would possess, for no one could excel in functioning as a human being if he

were liable to such excessive appetites and passions. The argument on

which Aristotle relied for this conclusion collapses, however, as we have

seen. Nothing in it, on its broad interpretation, excludes the possibility of

one�s acting unjustly from different appetites or passions, in particular,

appetites or passions that are consistent with one�s fulfilling one�s function

as a human being and achieving excellence in doing so, and any narrower

interpretation of the argument one gives so as to exclude this possibility

rests ultimately on an arbitrary preference for some forms of human func-

tioning over others. Hence, Aristotle�s naturalist program, too, falls short of

sustaining Plato�s idea that people whose lives were complete and fulfilling

would not be liable to themotives fromwhich unjust actions towards others

spring.

Nonetheless, it is important to note that the failure of Plato�s and

Aristotle�s programs to sustain this idea does not show that the idea itself

is bankrupt. To the contrary, the idea may still represent an important

insight into human nature. If it does, though, it may take empirical

studies of human lives to confirm this. Both Plato and Aristotle undertook

to establish its truth through a study of the very meaning of human


well-being. They undertook, that is, a quintessentially philosophical study.

The lesson of the failures of these studies to show that being just and honest

in one�s dealings with others is necessary to living a good life may well be

that a different kind of study is required to vindicate Plato�s idea. Its vindi-

cation, for instance, may require a study of the lives of particular people

who have lived well and whose success in life encompasses the develop-

ment and exercise of their distinctively human powers and capabilities.

Ideally, it would be a study that was designed to show whether such people

were significantly less susceptible to the kinds of motives from which

unjust and dishonest actions spring than others. But more realistically,

one might have to settle for a study designed to determine whether those

kinds of motive occurred less frequently in the lives of such people than in

the lives of people who have lived less well, either because of a lack in

development or exercise of their distinctively human powers and capabil-

ities or because of a lack of success in life despite the development and

exercise of these powers and capabilities. In any case, the studies would

follow the research protocols of the social sciences, and their result would

be the statistical generalizations characteristic of findings in those disci-

plines and not the universal and necessary truths that are the object of

philosophical study.

So we can be sure that such results would not satisfy Glaucon and

Adeimantus. They want Socrates to demonstrate a universal and necessary

incompatibility between being unjust and living a good life. Showing a

significant lack of susceptibility to the motives from which unjust actions

spring in those who live good lives could only seem to them a poor sub-

stitute. They have challenged Socrates to give a philosophical defense of

justice, and such a defense must deliver stronger results than generaliza-

tions representing susceptibilities and tendencies. Yet wemust consider the

possibility that their challenge asks too much from a philosophical study.

We must consider, that is, the possibility that the form of eudaimonism

Plato originated and Aristotle remodeled is not fully defensible within

philosophy. Perhaps, then, in the end, the best defense of their theory

follows Aristotle�s program up to the point of defining, by reference to the

special powers and capabilities of human beings, human life�s distinctive

form and subsequently characterizing the best life for human beings as one

that is complete and fulfilling by virtue of the realization of these powers

and capabilities. Having reached this point, however, this proposed defense

Eudaimonism 91

of their theory abandons the effort to demonstrate that possession and

exercise of the virtue of justice is necessary in all cases to living a good

human life and appeals instead to contingencies of human life that would

seem to make highly improbable the absence of this virtue in people who

have, through the realization of their special human powers and capabil-

ities, achieved a complete and fulfilling life. Such a defense is not, to be sure,

a solution to the core problem of Plato�s Republic. But it does offer some hope

of refuge from the anxieties about the truth of Thrasymachus� nihilism that

lead Glaucon and Adeimantus to press their challenge on Socrates.


4 Utilitarianism

1. Impartiality

Both egoism and eudaimonism share an outlook of self-concern. They both

identify the perspective from which a person judges what ought to be done

as that of someone concerned with how best to promote his own good. On

either theory, then, the highest good for a person is that person�s own good,

whether this be his own happiness or his own well-being. Hence, on either

theory, ethical considerations are understood to have the backing of reason

insofar as they help to advance this good.

The self-concerned outlook prevailed in ancient ethics, for eudaimonism

was its dominant theory. Modern ethics, by contrast, has been marked by a

shift away from this outlook. Eudaimonism no longer dominates the field,

andwhile egoismhas continued to have supporters throughout themodern

period, theories that presuppose a different outlook from that of self-

concern have eclipsed it. These later theories do not identify the perspective

from which a person judges what ought to be done as that of someone

concerned with how best to advance his own good. Nor do they explain

reason�s backing of ethical considerations by showing how those consider-

ations help to advance that good. To explain this backing modern moral

philosophers have supposed, instead, that ethical considerations speak to

some other element in human personality than concern about one�s own

good. Some philosophers have placed this element within the powers of

reason themselves and supposed that we have a special rational capacity for

knowing our duty. Such philosophers typically support deontological theo-

ries of ethics. Others have pointed to our benevolent dispositions toward

our fellow man and other members of the animal kingdom as the element

towhich ethical considerations speak. These philosophers typically endorse

utilitarianism, which is the theory that, in its classical form, takes the

highest good to be the good of humankind or the good of sentient animals

93

generally. A third group synthesizes the chief themes of the first two.

Specifically, they endorse utilitarianism on the grounds that reason dictates

the pursuit of the good of humankind or the good of sentient animals

generally. Thus, like the first group, these philosophers also identify reason

as the element in human personality to which ethical considerations speak.

Utilitarianism is the subject of the present chapter.1 We will turn to deon-

tological theories in the following chapter.

Utilitarianism, like egoism and eudaimonism, is a teleological theory.

The prescriptions it grounds are judgments of what one ought to do in the

sense of what one would be well-advised to do in view of one�s ends and

interests. The end in view of which utilitarianism grounds the prescriptions

of ethics is the general good, understood either as the good of humankind

or, more comprehensively, the good of all animals capable of experiences

with which human beings can sympathize. To simplify our exposition,

though, let us assume the less comprehensive of these ends as the one on

which utilitarianism grounds the prescriptions of ethics. The good of

humankind, then, on our simplified exposition of utilitarianism, is the

ultimate end of ethics and thus the highest good for human beings. Acts

are right or wrong according as they further or impede the achievement of

this end. To be more exact, on the most careful formulation of the utilitar-

ian standard of right andwrong, an act is right if it furthers the achievement

of this end as well as, if not better than, any other act one could do in the

circumstances one faces. An act is wrong if it fails to further the achieve-

ment of this end or furthers it less well than some other act one could do in

those circumstances. This standard is commonly known as the Principle of

Utility or, in keeping with the utilitarian tradition that looks to Jeremy

Bentham as its founder and John Stuart Mill as its most eminent defender,

the Greatest Happiness Principle. On a formulation of the principle as a

1 The division of utilitarian theorists into two groups, those who identify general benev-

olence as the element in human personality to which ethical considerations speak and

those who identify reason as that element, captures a trend in the development of

utilitarian ethics. Classical utilitarians like Mill tend to belong to the first group. Since

Henry Sidgwick (1838–1900), who originated and developed, in his masterwork, The

Methods of Ethics, 7th edn. (London:Macmillan and Co., 1907), the idea of synthesizing the

chief themes of rationalist deontology and classical utilitarianism, utilitarians have

gravitated toward the second. The exposition of utilitarianism in this chapter is consis-

tent with the views of either group.


prescription, in accordance with this Benthamite tradition, it runs: So act as

to bring about as much happiness in the world as you can in the circumstances you face.

Bentham�s formulation of the Principle of Utility as a prescription about

the promotion of happiness reflects his acceptance of hedonism as the

correct theory of well-being. The formulation is thus characteristic of clas-

sical utilitarian theory. But it is not required of all versions of the theory.

The principle, in other words, is not wedded to hedonism. One could also

formulate it in a way that reflected acceptance of perfectionism as the

correct theory of well-being. Or one could formulate it in a way that

reflected acceptance of some mixed theory. To put the point generally, the

Principle of Utility does not presuppose any specific theory of either well-

being or happiness. It prescribes that one act so as to bring about as much

good overall in the world as one can, and how one conceives of that good,

whether one conceives of it as happiness or well-being, and whether one

defines it hedonistically, perfectionistically, or as some combination of the

two, determines a specific version of utilitarianism but is otherwise

unnecessary to its study. We can therefore, in pursuing the study, leave

aside questions about the substantive nature of this good.

At the same time, we do need to specify the procedure by which the

overall good an act brings about is calculated. Following Bentham�s model

of a hedonic calculus, let us take this procedure to consist in one�s first

assessing the amounts of good and evil the act brings about in the lives of

everyone it affects and then separately totaling all the good it brings about

and all the evil. Finally, one subtracts the latter total from the former. One

thus arrives at the net balance of good over evil that the act produces. If one

defines the overall good hedonistically, then the net balance of good over

evil one arrives at results from subtracting the total amount of pain the act

produces from the total amount of pleasure. And if one defines the overall

good perfectionistically, then the net balance of good over evil one arrives at

results from subtracting the degree and extent of the deficiencies in human

activity the act produces from the degree and extent of the excellences in

human activity it produces. To apply the Principle of Utility, then, one

canvasses the different actions a person can take in the circumstances he

faces, calculates for each the projected net balance of good over evil it would

produce, and then identifies the right action, the action the person ought to

do, as the one that will produce the greatest net balance of good over evil.

Accordingly, one could state the Principle of Utility as prescribing that one

Utilitarianism 95

so act, in the circumstances one faces, as to produce as great a net balance of

good over evil as one can.

On Bentham�s hedonic calculus, experiences of pleasure and pain figure

in the overall good an act brings about according to how intense they are

and how long they last. That is, one measures how much pleasure or how

much pain an experience of either consists of along two dimensions of the

experience, intensity and duration, and the resultant quantity of pleasure

or quantity of pain one attributes to the experience translates directly into

the amount of good or evil the experience represents. Thus, it is implicit in

Bentham�s calculus that experiences of pleasure and pain contribute,

positively or negatively, to the overall good an act brings about regardless

of whose experiences they are. Equally intense and long-lasting experien-

ces of pleasure contribute equally to the overall good whether they are

yours,mine, or some stranger�s, and similarly for equally intense and long-

lasting experiences of pain. The calculus is therefore, in a word, impartial

among all of the people whose experiences of pleasure and pain figure in

the calculation of the overall good that an act brings about. And this

impartiality is implicit in every utilitarian procedure for calculating this

good.

Thus utilitarianism, by virtue of this procedure, entails an impartial

outlook toward all of humankind. For the procedure requires that you

consider, when calculating the overall good that an act brings about, every-

onewhose life the act affects for good or ill and include in the calculation all

benefits and harms each receives or will receive as a result of the act.

Further, it requires that you treat a benefit or a harm to someone as adding

or subtracting the same amount of good to the overall good the act brings

about regardless of who the person benefited or harmed is. In particular, it

does not matter whether this person is yourself, a friend, or a stranger.

Clearly, then, utilitarianism rejects the outlook of self-concern common to

egoism and eudaimonism. It rejects as well outlooks that are likewise

partial to some people or biased against others. It therefore rejects clannish

and chauvinistic outlooks and, indeed, any outlook that involves favoritism

toward one group over another. To act as the Principle of Utility prescribes

means that one regard the good of others as equally important as one�s own

good or the good of one�s friends, family, countrymen, or the like.

Bentham�s dictum �Everyone to count for one, no one to count for more

than one� expresses this utilitarian ideal of impartiality succinctly.


The ideal, as we�ve seen, is inseparable from the utilitarian procedure for

calculating the overall good that an act brings about. Yet it is not the sole

ideal that utilitarianism endorses. Since utilitarianism�s orientation is

toward benefiting humankind, we should understand its ideal of impartial-

ity as belonging to the grander ideal of a life devoted to making the world a

better place. Moreover, �being,� in Mill�s words, �a direct emanation from

[the Principle of Utility],� the ideal of impartiality belongs to this grander

ideal as an integral part and not merely as an add-on.2 It emanates directly

from the Principle of Utility, for to take the good of humankind as the

highest good means treating all parts of that good, which is to say, the

well-being of each and every human being, as equally important. Failure

to do so would make about as much sense as treating one�s well-being on

Tuesdays and Thursdays as more important to advancing one�s own good

than one�s well-being on the other days of the week. Differential treatment

in either case, whether it is favoritism toward some people because of who

they are or favoritism toward some times in one�s life because of the days of

the week in which they occur, is wholly arbitrary. And to incorporate it into

one�s procedure for determining how best to advance the good of human-

kind or one�s own good, as the case may be, is likely to make things worse,

not better. Hence, the complete ideal of human life that utilitarianism

endorses is that of a life of impartial service to making the world a better

place.

This is, without question, a morally admirable ideal. Its similarity to the

Christian ideals of universal love and unconditional charity toward all is

unmistakable. �In the golden rule of Jesus of Nazareth,�Mill wrote, �we read

the complete spirit of the ethics of utility. To do as you would be done by,

and to love your neighbor as yourself, constitute the ideal perfection of

utilitarian morality.�3 Jesus, as we know, set very demanding ideals for

human beings, and the utilitarian ideal is no less demanding. The question,

then, is how one could realize it in a human life. The difficulty of living up to

Christian ideals is a central theme of Christianity and one of the chief

lessons of the religion�s teachings. Much, in fact, has been written in

criticism as well as in praise of Christian ethics on account of this theme.

The difficulties of realizing the utilitarian ideal are, if anything, even more

pronounced and similarly the basis of much criticism.

2 Mill, Utilitarianism, ch. 5, par. 36. 3 Ibid., ch. 2, par. 16.

Utilitarianism 97

2. Two problems

One immediate difficulty of realizing the utilitarian ideal is due to the very

impartiality that is built into the procedure for applying the Principle of

Utility and that separates utilitarianism from egoism and eudaimonism.

This impartiality, as we noted, consists in treating everyone�s interests as

equally important and so, in particular, in treating one�s own interests or

the interests of one�s friends and close kin as no more important than the

interests of mere acquaintances and strangers. And while the idea of treat-

ing one�s own interests as no more important than the interests of a

stranger has obvious moral appeal in the humility and solidarity with

humanity it expresses, the idea of regarding the interests of one�s friends

as no more important than the interests of some stranger hardly seems

similarly inspiring. Quite the contrary, it seems to contradict what it means

to be someone�s friend. By the same token, if you found someone who

regularly treated the interests of his or her spouse as no more important

than those of a stranger, you would ordinarily take this as a sign of a

marriage gone bad and not of a person exercising moral virtue. Defenders

of utilitarianismwhohave bitten the bullet, so to speak, and declared, as the

early utilitarianWilliam Godwin (1756–1836) did, that if he were to happen

upon a burning building in which an important public official and his valet

were trapped, he would first save the official even if the valet were his son,

have simply seemed out of touch with human life rather than wise in the

ways of morality.

This is not, of course, to say that there are no circumstances in which

doing the right thing requires treating the interests of one�s friends, one�s

spouse, or one�s children as no more important than the interests of mere

acquaintances and strangers. One who has assumed the responsibilities of a

judge, a magistrate, or an administrator of government services, for exam-

ple, and acts rightly in carrying them out, acts impartially, and this means

that he or she, in fulfilling these responsibilities, treats the interests of

friends and close kin as no more important than the interests of mere

acquaintances and strangers. But we expect such impartiality of judges,

magistrates, and government administrators because of the public offices

they hold. We do not similarly expect it of private individuals as they go

about their daily lives. Utilitarianism, however, by making such impartial-

ity the rule in everyday decisions about what to do would appear to expect


people always to act as if they were carrying out the responsibilities of a

public official. And such expectations, if they truly represent the perspec-

tive on life utilitarianism recommends, threaten the theory�s credibility.

A second difficulty of realizing the utilitarian ideal arises when one tries

to fit adherence to standards of justice and honesty into a life lived in

pursuit of this ideal. Admittedly, this might not at first seem to be much

of a difficulty. After all, just and honest actions are actions that typically

promote the well-being of others. Indeed, this feature of them is what led

Thrasymachus to dismiss as foolish and weak people who adhered to stand-

ards of justice and honesty when they could ignore them with impunity.

Just and honest actions therefore, because they seem to be actions contri-

buting to the good of humankind by virtue of serving the interests of those

on their receiving end, may seem to fit easily into a life lived in pursuit of

the utilitarian ideal. Yet problems emerge nevertheless when one tries to

square adherence to these standards with faithfully following the Principle

of Utility in the conduct of one�s life. These problems emerge because in

some situations, so it seems, one can bring about a greater net balance of

good over evil by acting unjustly or dishonestly than one could bring about

through just and honest actions.

Our example of finding a lost purse containing a huge wad of cash nicely

illustrates this point. Although you might initially think that to follow the

Principle of Utility in this situation required returning the purse along with

the cash to its owner, a little reflection should lead you to the opposite

conclusion. For you need to consider, when applying the Principle of Utility,

every act you could do in this situation, and while more overall good might

come from returning the purse and cash to their owner than keeping the

cash for yourself and discarding the purse, there are doubtless other acts

you could do that would bring about even more overall good than either of

these. A hugewad of cash, after all, can be used to do a lot of good for people

who urgently need help or whose lives are so impoverished as to lack many

of the basic necessities of a decent life. So to apply the Principle of Utility

you would need to consider whether giving the cash to some charitable

organization whose programs help such people would in fact bring about

more overall good than returning it to its owner.

To get down to cases, suppose it is evident from your examining the

contents of the purse that the owner is a well-to-do dentist who lives in a

fashionable suburb, or suppose on looking at her driver�s license you

Utilitarianism 99

recognize her name as that of a prominent and wealthy member of the

community, you should then confidently conclude that the loss she will

suffer, if you do not return the cash to her, will be more than offset by the

great good that would come from giving it to Save the Children or Second

Harvest. Calculatingwith respect to eachof these acts the good and evil that it

is likely to bring about should lead you to see that the net balance of good

over evil of the former is significantly less than that of the latter. Following

the Principle of Utility in this situation appears, then, to require that you act

dishonestly. The Principle, in other words, when applied to the situation,

yields a prescription that directly conflicts with what basic standards of

honesty prescribe. Such a conflict is plainly a serious problem for the theory.

The problem, moreover, goes beyond the conflict. For the example also

shows that the theory treats considerations of justice and honesty in a way

that is foreign to even an elementary understanding of how these consider-

ations figure in deliberations about what to do. When you find a lost purse

containing a hugewad of cash, you come into possession of another person�s

property without their consent. Recognizing that the purse and its contents

belong to someone else, you realize, even on an elementary understanding of

property, that you are not free to dispose of them as you see fit, that to do so

would be to transgress the limits that another�s being the owner of this purse

places on your conduct, and that such a transgression would be an act of

dishonesty. Yet this fact about the purse and its practical implications do not

seem to register anywhere in the procedure for applying the Principle of

Utility to the situation you face. For on this procedure, you regard every act

that you can do in the situation as an option until you determine that it will

bring about less good overall than some other act you can do. Consequently,

until you make such a determination about some act, you are to regard

yourself as free to do it. That the purse is not yours but someone else�s does

not then, it would seem, enter into your thoughts as restricting what actions

youmay consider doing. The dishonesty of disposing of the cash in the purse

without its owner�s consent would therefore appear, on this procedure for

applying the Principle of Utility, never in itself to figure in your deliberations

as a consideration that bears onwhether the act is right or wrong. Hence, the

conflict between theprescription that results fromapplying the Principle and

the prescription that basic standards of honesty yield signifies an evendeeper

problem in the theory than that its prescriptions can conflict with those that

basic standards of justice and honesty yield.


3. Consequentialism

The source of this problem is utilitarianism�s supposition that at the foun-

dation of ethics is a single standard of right and wrong, the Principle of

Utility or Greatest Happiness Principle, that prescribes optimal promotion

of the highest good. As a result of this supposition, the theory defines right

action strictly by its consequences for good or ill in comparison with the

consequences for good or ill of the other actions the agent can do in the

circumstances, and similarly for its definition of wrong action. In other

words, according to utilitarianism, the sole determinants of whether an

act is right or wrong are the consequences of that action for good or ill in

comparison with the consequences of the other actions available to the

agent in the circumstances of his action, and therefore the only consider-

ations that bear on whether an action is right or wrong are these conse-

quences. Considerations of honesty and justice, by contrast, encompass

other things besides the consequences of actions for good or ill. They

encompass as well relations to past actions and present circumstances.

The dishonesty of spending or giving away cash that one knows belongs

to someone else and that one possesses without its owner�s consent is not a

matter of the consequences for good or ill of the action. It is rather a matter

of the action�s circumstances (viz., that the cash belongs to someone else)

and the absence of certain past actions (viz., that the owner never consented

to one�s disposing of her cash as one sees fit). Similarly, the dishonesty of

telling a lie is not a matter of the consequences of saying something one

knows to be false to someone but rather amatter of circumstances (viz., that

what one said to this person is false and that one said it with the intention of

deceiving him). A theory that excludes such factors from being determi-

nants of right and wrong actions, a theory that restricts the determinants of

right and wrong actions to an action�s consequences either absolutely or in

comparison with the consequences of other actions, exemplifies conse-

quentialism. And the difficulty for utilitarianism that our example of your

finding a lost purse containing a huge wad of cash illustrates is due to the

consequentialism of the theory.

Standards of justice and honesty includemany of the basic rules of social

conduct, rules that prohibit or require certain conduct because of present

circumstances or past actions and that we learn at an early age. Among

them are the rules that prohibit our taking what belongs to others without

Utilitarianism 101

their consent and that require our telling the truth. At first, because chil-

dren have limited experience of the social world and abbreviated develop-

ment of their capacities for self-restraint, these rules are taught as

unbreakable. In time, however, as children venture out from their homes

and families into the wider social world of schools, churches, clubs, work,

and community affairs, they acquire a more complex understanding of

these rules and greater confidence in their judgment about how to apply

them. The rigidity of our early obedience to them relaxes somewhat as we

learn about exceptions and about how to incorporate considerations of the

consequences of our actions into our judgments about how the rules apply

to our circumstances. Sometimes, as we learn, the consequences of obedi-

ence to one of these rules may be so bad that avoiding those consequences

justifies our disobeying the rule: when emergencies occur – when the

neighbor in the apartment across the hall slices off his thumbwith a carving

knife and you are his only hope for getting quickly to the hospital – your

breaking a promise to visit a lonely aunt or your taking your roommate�s car

without his permission is justified. Nonetheless, we still understand the

rules of justice and honesty as placing limits on our conduct, even though

we also understand that those limits can be justifiably crossed in extreme

circumstances. That is, leaving aside such emergencies and other crises, we

continue to understand the rules as excluding from the range of actions

open to us actions that would break the rules and so to regard these actions

as not among our options, notwithstanding their having better consequen-

ces than the consequences that will result from our obeying the rules.

Accordingly, we continue to understand the rules as binding upon us.

Consequentialism, in contrast, abandons this understanding of the rules

of justice and honesty as binding. It invites us not only to take the conse-

quences of an action as the sole determinants of whether it is right or wrong

but also to see every action we can perform as, in principle, among our

options until such time as it is determined that its consequences would be

worse than the consequences of some other action open to us. It thus

opposes directly our common understanding of the rules of justice and

honesty as excluding from the range of actions open to us those that

would break these rules. According to consequentialism, the rules of justice

and honesty do not set barriers to our conduct but rather represent good,

though less than surefire guides to what actions will have the best conse-

quences in the circumstanceswe face. Definitions of �right� and �wrong� that


exemplify consequentialism remove from those words the import they

ordinarily have as terms for actions that respect or violate limits on conduct

that moral standards impose, and this loss is most apparent in the way

theories that incorporate such definitions render the rules of justice and

honesty.

4. Mill�s restatement of utilitarianism

Mill was keenly aware of these problems. The second, in particular, was the

source of the popular criticism of utilitarianism that it was a philosophy of

expediency and as such the enemy of justice. The criticism seemed well

targeted when leveled at Bentham�s statement of the theory, andMill, in his

famous short work, Utilitarianism, set out to restate Bentham�s ideas so as to

avoid this criticism. Specifically, he recognized the embarrassing prescrip-

tions the theory yields when one takes its supposition that the Principle of

Utility is the foundation of ethics to mean that one should apply the

Principle directly to one�s circumstances when determining what one

ought to do. To avoid these embarrassing results, then, it is necessary to

deny that utilitarianism requires that one apply the Principle directly to

one�s circumstances. To say that it is the fundamental principle of morality,

Mill argued, is not to say that the theory recognizes no other moral princi-

ples. �It is a strange notion,� he wrote, �that the acknowledgement of a first

principle is inconsistent with the admission of secondary ones.�4 To the

contrary, like other systems of ethics, utilitarianism, too, recognizes secon-

dary principles whose use is necessary to achieving the end that the funda-

mental principle prescribes. �Whatever we adopt as the fundamental

principle of morality,� Mill declared, �we require subordinate principles

to apply it by; the impossibility of doing without them being common to all

systems, can afford no argument against any one in particular.�5

Hence, on Mill�s restatement of utilitarianism, the theory recognizes a

plurality of moral principles that are subordinate to the Principle of Utility.

They are subordinate in the sense that their validity is due to their impor-

tance to people�s achieving the end that the Principle of Utility prescribes,

to their achieving, that is, the greatest overall good for humankind. To

achieve this end, Mill held, it is necessary for people to accept and follow

4 Ibid., ch. 2, par. 24. 5 Ibid.

Utilitarianism 103

these subordinate principles. Accordingly, he refers to them as �corollaries

from the Principle of Utility� and remarks that they represent the collective

wisdom of humankind concerning what actions best promote its end.6

�There is no difficulty,� Mill observed, �in proving any ethical standard

whatever to work ill if we suppose universal idiocy to be conjoined with

it; but on any hypothesis short of that mankind must by this time have

acquired positive beliefs as to the effects of some actions on their happiness;

and the beliefs which have thus come down are the rules of morality for the

multitude.�7 On Mill�s theory, then, people, when determining what they

ought to do, should consult these rules of morality and apply them directly

to their circumstances, for by doing so they will best fulfill morality�s

fundamental principle. Of course, Mill conceded, these rules of morality

are not immutable. They change as the conditions of human life change and

as people gain greater wisdom about the world and how it works. �But to

consider the rules of morality as improvable,� Mill cautioned, �is one thing;

to pass over the intermediate generalization[s] entirely and endeavor to test

each individual action directly by the first principle is another,�8 and util-

itarianism�s critics, he maintained, have misrepresented the theory in sup-

posing that it requires the latter.

Mill�s restatement of utilitarianism thus removes the Principle of Utility

from directly determining whether an act is right or wrong. As a result, the

version of the theory Mill presented does not define right action strictly by

the action�s consequences for good or ill in comparisonwith the consequen-

ces for good or ill of the other actions that the agent can perform in the

circumstances he faces; and similarly for its definition of wrong action.

Instead, the theory appeals to principles subordinate to the Principle of

Utility, the rules of morality, in defining right and wrong. That is, on Mill�s

version of the theory, an action is right if it conforms to the rules ofmorality

and wrong if it violates them, where the rules of morality are those princi-

ples that wisdom has shown to be the principles human beings must

observe and follow if they are to secure and promote the general good.

Since the basic standards of justice and honesty are among these principles,

this version of the theory does not yield prescriptions that conflict with

what the basic standards of justice and honesty prescribe. It does not, in

particular, tell you that the right thing to do when you find a lost purse

6 Ibid. 7 Ibid. 8 Ibid.


containing a hugewad of cash is to give the cash to a charitable organization

that will use it to bring about more good for humankind overall than could

possibly result from returning the cash to its owner. Rather it tells you to

consult the rules of morality and apply them directly to the situation, and

since the basic standards of justice and honesty are the applicable rules in

these circumstances, it tells you that the right thing to do is to return the

purse and the cash to its owner. By interposing a set of secondary principles

between the primary principle, the Principle of Utility, and the circum-

stances of action Mill was therefore able to formulate a version of utilita-

rianism on which neither right nor wrong action is determined by direct

calculation of the action�s consequences and those of its alternatives in the

manner that Bentham had prescribed.

Yet Mill was careful, in presenting this version of utilitarianism, not to

deny that the consequences of an action can ever determine its moral quality

or that circumstances can arise inwhich, because of grave consequences that

following somemoral rule would have, the right thing to do is to act contrary

to it. Even the rule requiring that one tell the truth, Mill observed, �admits of

possible exceptions . . . the chief of which is when the withholding of some

fact (as of information from a malefactor, or of bad news from a person

dangerously ill) would save an individual (especially an individual other

than oneself) from great and unmerited evil, and when the withholding can

only be effected by denial.�9 Further,Mill proposed, these exceptions, like the

rules themselves, are generalizations whose validity is due to the importance

to achieving the general good for humankind of their being recognized and

followed. The samewisdom, that is, that has shownwhat rules human beings

must follow to achieve the end that the Principle of Utility prescribes also

teaches us about the special circumstances inwhich onemust set a rule aside

for the same purpose. The Principle of Utility, in other words, on Mill�s

restatement of utilitarianism, not only determines the rules of morality but

carves out their exceptions as well. These exceptions are therefore part of the

system of rules by which the Principle of Utility is applied, just as tax

exemptions are part of the tax code. And by understanding exceptions to

the rules of morality in this way Mill was able to acknowledge them consis-

tently with his program of removing the Principle of Utility from directly

determining whether an act is right or wrong.

9 Ibid., ch. 2, par. 23.

Utilitarianism 105

In fact, according to Mill, the only times when one must consult the

Principle of Utility directly to determinewhat act ought to be done arewhen

two or more of the secondary principles apply to one�s circumstances and

yield conflicting prescriptions.10 Suppose, for example, that you have been

taken into another�s confidence and have given him assurances that you

will keep his secret. You would then find yourself in conflictual circum-

stances of the sort Mill had in mind if someone were to ask you a direct

question to which you could not respond truthfully without betraying that

confidence. In these circumstances, some method of resolving the conflict,

some method of determining whether the right thing to do is to keep the

secret or respond truthfully, is necessary, and one of the virtues of utilita-

rianism,Mill argued, as comparedwith the traditional systems of ethics that

recognize a plurality of moral rules but treat each as having independent

authority, is that it contains a method for resolving these conflicts. Its

method consists in one�s directly consulting the Principle of Utility to

determine which act of complying with the rules best achieves the end on

which they are all founded, for by determining which act best achieves this

end, one determines which rule should have precedence in the circum-

stances one faces. In other words, although one directly consults the

Principle of Utility to determine what act ought to be done, the act�s being

right is still a matter of its being prescribed by the rules of morality. Hence,

even in these circumstances, Mill�s endorsement of direct appeal to the

Principle of Utility is fully compatible with his thesis that appeal to sub-

ordinate principles is necessary to determining whether an act is right or

wrong. His acknowledgment of the propriety of directly appealing to the

fundamental principle of morality to resolve conflicts between prescrip-

tions that can arise when two or more of the secondary principles apply to

one�s circumstances is therefore consistent with his denying that one ever

determines, on the utilitarian system, whether an act is right or wrong by

applying the Principle of Utility directly, which is to say, independently of

any of the principles subordinate to it.

Mill�s argument shows, then, how one can, by restating utilitarianism so

that the Principle of Utility never decides questions of right and wrong

independently of its subordinate principles, assert the Principle as the

fundamental principle of morality without being committed to the

10 Ibid., ch. 2, par. 25.


embarrassing prescriptions that threaten the credibility of the theory as

Bentham stated it. In addition, it articulates a version of utilitarianism that

escapes the first problem that we noted in Bentham�s version, the problem

of misplaced impartiality. By limiting the role of the Principle of Utility so

that it does not decide questions of right and wrong independently of other

principles, Mill�s version avoids the necessity of always treating everyone�s

interests, whether they are the interests of close friends or complete strang-

ers, as equally important when determining what one ought to do. Such

impartiality, as we saw, is inherent in Bentham�s version by virtue of its

being built into his procedure for applying the Principle of Utility to the

circumstances of action. Hence, the problem arises for this version because

questions of right and wrong, on Bentham�s statement of the theory, are

decided in every case by directly applying the Principle of Utility in accord-

ance with this procedure. On Mill�s restatement of utilitarianism, by con-

trast, impartiality is a feature of the subordinate principles and their

exceptions in virtue of their being corollaries to the Principle of Utility.

Accordingly, one acts consistently with the utilitarian ideal and, in partic-

ular, the ideal of impartiality it implies so long as one�s actions conform to

these principles as qualified by their exceptions. And just as the subordinate

principles include basic standards of justice and honesty, they include as

well principles that make possible friendship and close ties among the

members of a family, for such personal relations are among the most

important elements of human well-being. That is, the principles permit

one, as one goes about one�s daily life, to treat the interests of friends and

family as more important than those of mere acquaintances and strangers

insofar as such partiality is necessary to sustaining these personal relations.

On Mill�s version of utilitarianism, therefore, favoring one�s friends and

loved ones, because their interests matter to one more than the interests of

acquaintances and strangers, need not and typically does not put one at

odds with the utilitarian ideal.

5. An inconsistency in Mill�s restatement

Mill�s restatement of utilitarianism shows how one can formulate the

theory to avoid the embarrassing prescriptions to act unjustly and dishon-

estly that Bentham�s statement of it entails. Yet the question remains

whether it also preserves our understanding of the basic standards of justice

Utilitarianism 107

and honesty as placing limits on our conduct. How canMill�s explanation of

these standards as principles subordinate to the Principle of Utility square

with our understanding of them as binding? Plainly, it can do so only if one

can understandMill�s restatement as incorporating a suitable conception of

these principles. Not any conception will do. Mill himself seems not to have

recognized this, for he shifts in his several discussions of them between

different conceptions. In particular, while it is certain that he saw these

principles as indispensable to the utilitarian system of ethics, it is uncertain

whether he saw their indispensability as due exclusively to inescapable

limits on human powers of discernment, foresight, and calculation or also

to the necessity for the enjoyment of certain essentials of humanwell-being

of there being rules withwhich people generally comply. And depending on

which of theseways one sees their indispensability a different conception of

them is implied. Thus correcting Mill�s inconsistency in his discussion of

these subordinate principles will make clear which conception of them he

needs if his restatement of utilitarianism is to preserve our understanding

of the basic standards of justice and honesty as placing limits on our

conduct.

Mill�s inconsistency is due to his running together at least two different

kinds of rule that exist in human life. Some rules, such as �See your dentist

twice a year� and �Don�t put off till tomorrow what you can do today,�

embody lessons about avoiding trouble and achieving success. Others, such

as �Do not butt in line� and �Awedding ring is worn on the ring finger of the

left hand,� regulate conduct within practices and institutions like queues

and marriage. Games with a long tradition of play offer good examples of

both kinds. Consider chess. The rules that define how the game is played are

the ones you will naturally think of first. These are the rules that determine

where each of the pieces is placed at the beginning of the game, which

player moves first, how each piece can be moved, how it can be captured,

what counts as winning, and so forth. �No two pieces can occupy the same

square at the same time� is a rule that governs the movement of all pieces.

�Kings move one square at a time in any direction� is the rule for the

movement of kings. Further, there is a somewhat complicated rule for the

special coordinatemovement of kings and rooks called castling.11 Let us call

11 The rule is: If the king and either rook occupy the squares on which they were placed at

the beginning of the game and neither has ever been moved from that square and if


all such rules the institutional rules of chess. In addition, there are what I�ll

call strategic rules that good chess players know and follow. �Control the

center of the board� and �Castle early� are examples of strategic rules that

apply to the early stages of the game. Following the first of these improves

your chances of putting your opponent on the defensive and ultimately

checkmating your opponent�s king. Following the second makes your own

king less vulnerable to attack by your opponent�s pieces and so to being

checkmated. These strategic rules, in other words, embody lessons about

how to achieve victory and avoid defeat that experience at playing the game

has taught.

One important difference between these two kinds of rule is this.

Institutional rules regulate conduct within some practice or institution

categorically in the sense that any act that breaks an institutional rule is a

wrong act within the practice or institution of which the rule is a part. Take

games again as the paradigm of a practice. The institutional rules of a game

regulate its play categorically in the sense that any violation of them is

necessarily a wrongmove and as such is immediately nullified andmay also

be subject to a sanction. The game�s institutional rules thus place limits on

what a player can do in playing the game. They can be said, then, to be

binding. Strategic rules, by contrast, do not apply categorically. To break

one of them is not necessarily to do something wrong. Indeed, it may even

be the best act one could do in the situation. In a game, for instance, some-

times one makes a better move by breaking some strategic rule than by

following it. Strategic rules, in other words, do not place limits on what one

can do within the game. They are not binding. Rather they are guides that

good players generally follow because even the best players cannot calcu-

late all the possible consequences of each potential move they could make

at a given point in the game. And since the reason for following a strategic

rule is that experience has taught that doing so generally improves one�s

chances of winning, one has no reason to follow the rule in the occasional

situation in which one can see that making a move breaking the rule will

improve one�s chances of winning even more than if one made a move that

followed it. Bridge players who are familiar with the chapter �Defensive

there is no piece between themon the rank that contains the squares they occupy, then

one canmove the king two squares along that rank in the direction of the rook and then

move the rook to the square on the other side of the king that is next to the square the

king now occupies.

Utilitarianism 109

Rules – And When to Break Them� in the popular instructional guide

5 Weeks to Winning Bridge by Sheinwold will understand this point

immediately.12

Mill�s shift between different conceptions of the principles subordinate

to the Principle of Utility consists in his sometimes treating these principles

as strategic rules and sometimes treating them as institutional rules. When,

for example, Mill deals with the objection that utilitarianism is woefully

impractical because people generally lack the knowledge and cognitive

capacities necessary to do the calculations that following the Principle of

Utility requires, he treats them as strategic rules. They are, he says, like the

principles of any practical art, grounded in the experience of the art�s

practitioners and usable as guides that represent more reliable calculations

of what act will bring about themost good overall in the circumstances than

a fresh calculation made without regard to the experience of those who

have faced similar circumstances. On the other hand, when he deals with

the objection that utilitarianism is the enemy of justice because it ulti-

mately calls for expediency, he treats them as institutional rules. They

include, he says, the rules of morality, and it is in the nature of such rules

to define duties that are inescapable and backed by punitive sanctions. The

rules of justice, in particular, Mill maintains, establish a regime of individ-

ual rights in human affairs necessary to preserving peace among human

beings and developing their social feelings, and the great benefits of this

regime establish them as principles subordinate to the Principle of Utility

and justify their absolute obligation. �The moral rules which forbid man-

kind to hurt one another,� Mill wrote, � . . . are more vital to human well-

being than any maxims, however important, which only point out the best

mode of managing some department of human affairs. They have also the

peculiarity that they are the main element in determining the whole of the

social feelings of mankind. It is their observance which alone preserves

peace among human beings.�13

It should be clear that if Mill�s restatement of utilitarianism is to preserve

our understanding of the basic standards of justice and honesty as placing

limits on our conduct, then the principles subordinate to the Principle of

12 Alfred Sheinwold, 5 Weeks to Winning Bridge, rev. edn. (New York: Pocket Books, 1964),

pp. 388–406.13 Mill, Utilitarianism, ch. 5, par. 33.


Utility to which Mill appeals must be conceived of as institutional rules and

not as strategic ones. In other words, the way to fix Mill�s inconsistency is to

take as the correct conception of these principles the conception he

advanced in answering the objection to utilitarianism that it is the enemy

of justice. On this conception, because the principles subordinate to the

Principle of Utility are part of certain practices and institutions, because

they are institutional rules, they apply categorically to the conduct they

regulate. To break them is necessarily to do something wrong within the

practice or institution of which they are a part. It is to transgress the limits

they place on the conduct they regulate. And because the practices and

institutions of which they are a part are the right practices and institutions

for human beings collectively to have, as judged by the fundamental prin-

ciple of morality (i.e., the Principle of Utility), breaking these principles is

wrong absolutely. It therefore follows that if the basic standards of justice

and honesty can be shown to be part of practices and institutions that, as

judged by the Principle of Utility, are the right practices and institutions for

human beings collectively to have, which is to say, the practices and insti-

tutions the maintenance of which brings about the greatest good for

humankind, then Mill�s restatement will have succeeded in incorporating

the basic standards of justice and honesty into the utilitarian system of

ethics and at the same time preserved our understanding of these standards

as placing limits on conduct.

6. Rule utilitarianism

The version of utilitarianism that we have recovered fromMill�s restatement

of the theory is now commonly called rule utilitarianism. It is so-called because

on this version rules are what one evaluates by consulting the Principle of

Utility. Individual acts are then determined to be right or wrong by the rules

that, according to these evaluations, are the right rules for human beings to

follow. In contrast, the version that originates with Bentham is commonly

called act utilitarianism because on Bentham�s version one applies the

Principle of Utility directly to individual acts to determine whether they are

right or wrong. The advantages of rule utilitarianism over act utilitarianism

are, as we�ve seen, that rule utilitarianism avoids the embarrassing prescrip-

tions that act utilitarianismyields and that it squares with our understanding

of the basic standards of justice and honesty as placing limits on our conduct.

Utilitarianism 111

These advantages do not come without a cost, however. Rule utilitarianism,

by interposing a set of moral rules between the Principle of Utility and the

circumstances of action and making right and wrong action a matter of

compliancewith and violation of these rules, creates a problemof explaining

why it is reasonable for a person to act rightly in circumstances in which his

performing a different action would better promote the general good. It

creates a problem, that is, of explaining why it is reasonable to act rightly

in circumstances inwhich act utilitarianismwould determine that the action

was wrong. And the theory�s sharpest critics have exploited this problem in

pressing what is the most serious objection to the theory. It is an objection

that seems to put the theory in the untenable position of being either an

unsatisfactory formof utilitarianismor an unsatisfactory formof deontology.

The problem is evident when one considers someone who faces circum-

stances in which, to comply with the secondary moral rules, he must

forbear from doing the act that in those circumstances would best promote

the general good. Our example of your finding a purse containing a huge

wad of cash suggests such circumstances, for one can easily imagine how in

these circumstances your returning the purse and the cash to their owner

would bring about less good in the world than your donating the cash to

someworthy charity. Suppose, then, that this is so and that secondarymoral

rules nonetheless require you to return the purse and the cash to their

owner. According to rule utilitarianism, therefore, you act rightly if you

do so and wrongly if you do not. You act rightly because returning the purse

and the cash to their owner is required by the right rules, and they are the

right rules because general compliance with them is more conducive to

promoting the good of humankind than general compliance with any of

their alternatives. So it is natural to think that the point of following these

rules is to promote the general good. Yet if this were the point of following

the rules, then you would be choosing to comply with them for the purpose

of promoting the good of humankind in circumstances in which you could

better fulfill this purpose by violating them. You would have more reason,

in other words, to violate the rules by donating the cash to a worthy charity

than to comply with them by returning the purse and the cash to their

owner. Hence, the question: why would it be reasonable to comply with the

rules in such circumstances? And failure to find a satisfactory answer would

mean that the theory could not explain why, on its account of right and

wrong action, it was always reasonable to act rightly.


The problem, it is important to realize, is not with Mill�s observation

about the necessity of certain practices and institutions for the enjoyment

of peace, security, individual liberty, and other essentials of human well-

being. Nor is it with the companion thesis about the necessity of general

acceptance of and compliance with the rules of these crucial practices and

institutions for the maintenance of their effectiveness. Bentham, too, rec-

ognized that such practices and institutions must be in place in human

society and their rules generally accepted and followed if human beings are

to have tolerably decent lives. These observations are commonplace in

utilitarianism. Consequently, no form of the theory, neither act utilitarian-

ism nor rule utilitarianism, treats violations of the rules these crucial

practices and institutions comprise lightly. On either form, because com-

pliance upholds these practices and institutions and violation undermines

them, the presumption in any situation will be that one acts rightly in

following the rules and wrongly in breaking them. The reason should be

clear. To weaken the authority of the practices and institutions is to jeop-

ardize people�s enjoyment of the essential human goods they make possi-

ble, and a violation of the rules of these practices and institutions weakens

their authority inasmuch as it weakens the violator�s own disposition to

follow the rules and, if made known to others, has a similar effect on them.

Hence, in view of the threat to the general good that violation of the rules

creates, the presumption that such violation is wrong is substantial.

On act utilitarianism, this presumption is reflected in the weight that is

given, in the procedure for applying the Principle of Utility, to harms that

would result from weakening the authority of these practices and institu-

tions. Accordingly, for a violation of their rules to be justified the good that

it brings about must not only be so great as to overmatch whatever distinct

harms it causes specific individuals to suffer but also be so great as to

overmatch the presumed harms that result from its weakening this author-

ity. On rule utilitarianism, by contrast, the presumption is reflected in the

inviolability of the rules in all circumstances except for those in which a

violation is necessary in order to comply with some other rule. Accordingly,

for the violation to be justified in such circumstances the rule with which

one complies must take precedence over the rule one violates, where

precedence is determined by directly applying the Principle of Utility to

those circumstances. The difference, therefore, between act utilitarianism

and rule utilitarianism does not consist in whether the theory recognizes

Utilitarianism 113

practices and institutions whose existence is crucial to human beings�

enjoying a tolerably decent life. Both recognize such practices and recog-

nize, too, the importance of people�s generally accepting and following the

rules the practices comprise. The difference consists, rather, in how those

rules figure in the theory�s procedure for determining whether an act is

right or wrong. In act utilitarianism they figure as factors in the calculation

of the overall good an act could bring about in the circumstances one faces,

factors that have weight in view of the harms that violation of these rules

would cause. In rule utilitarianism, they figure as authoritative constraints

on the range of actions that one is permitted to do in those circumstances.

And it is their figuring as authoritative constraints that generates the prob-

lem the theory has in explaining why it is reasonable to comply with these

rules when one could promote greater good overall for human beings by

violating them.

It generates the problem because, as authoritative constraints, the rules

override all other factors one might consider in determining what it is right

to do. Their authority, in other words, excludes consideration of any act

outside ofwhat the rules permit. In particular, it excludes any act thatmight

bring about more good overall for human beings than any of the acts the

rules permit. To do such an act would be to act against the authority of the

rules, and on the rule utilitarian theory such disobedience is the mark of

wrongdoing. At the same time, the rules derive their authority solely from

the Principle of Utility. What makes them authoritative, in other words, is

that general acceptance of and compliance with them is more conducive to

promoting the good of humankind than general acceptance of and compli-

ance with any of the alternatives. But if the ultimate authority is the

Principle of Utility, if what makes its subordinate rules authoritative is

their being conducive to promoting the general good, then it is hard to

see on what grounds an act that best realizes the Principle, which is to say,

an act that brings about themost good overall for humankind, could ever be

understood as disobedient of that authority. If the sole basis for the author-

ity of the rules is their conduciveness to promoting the general good and

being subject to their authority is what makes it reasonable to follow them,

then it is hard to see why one should regard them as authoritative and so

reasonable to follow in circumstances in which breaking them would be

more conducive to promoting the general good. Obedience to them in such

circumstances appears merely to be blind.


In light of this problem, it is hard to resist the conclusion that rule

utilitarianism is an unsustainable compromise. Insofar as it aims to capture

the understanding of moral rules as authoritative constraints on conduct

that is characteristic of deontology, it fails to give an adequate account of

that authority. And insofar as it aims to adhere to the logic of utilitarianism,

it falls short of being completely faithful to that logic.

7. Act utilitarianism revisited

Might there be a way of saving act utilitarianism from the embarrassing

prescriptions it yields? There is no way of excising these prescriptions from

the theory, to be sure. But perhaps one can sideline them instead. The most

sophisticated defenses of act utilitarianism take this tack. The cornerstone

of these defenses is the distinction between the evaluation of an action as

right or wrong and the evaluation of the motives behind an action as

sterling or corrupt. Defenders of utilitarianism from Mill forward have

invoked this distinction to correct common mistakes about the theory

that result from failure to appreciate its consequentialist character. �He

who saves a fellow creature from drowning,� Mill wrote, �does what is

morally right, whether his motive be duty or the hope of being paid for

his trouble.�14 The action�s being right, Mill is saying, is a separate matter

from its being done from a sterling or corrupt motive. And conversely an

action�s being done from a sterlingmotive is no guarantee of its being right;

nor would its being done from a corrupt motive guarantee that it would be

wrong. The point is well made in the old chestnut about the road to hell

being paved with good intentions. With this point in mind and bearing in

mind as well that act utilitarianism identifies acting rightly with acting in

accordance with the Principle of Utility, it should be clear that someone

who can be relied on to act rightly is not necessarily someone who consis-

tently makes efforts to follow the Principle of Utility. He may, instead, act

from different motives. More exactly, his personality or character may

reflect values and goals whose achievement requires his following princi-

ples other than the Principle of Utility. In short, even though act utilitari-

anism identifies acting rightly with acting in accordance with the Principle

of Utility, to be a reliable performer of right actions may nonetheless

14 Ibid., ch. 2, par. 18.

Utilitarianism 115

require one�s following principles other than the Principle of Utility. And

because of this possibility, the embarrassing prescriptions that result from

applying the Principle of Utility to situations like that of finding a lost purse

containing a huge wad of cash may not be as problematic for act utilita-

rianism as they initially seem. Or so the sophisticated defenders of the

theory whose argument we are now considering maintain.

Their argument consists of several steps. Thefirst is to separate, aswehave

just done, questions about what are the right actions to do from questions

about what are the right motives from which to act. While the theory�s

answers to questions of the former kind are in each case the action that

would conform to the Principle of Utility in the circumstances the agent

faces, the theory�s answers to questions of the latter kind are the motives

indicative of those traits of character that make a person the most reliable

performer of right actions that he could be. The argument�s second step,

then, is to point out, in view of this separation of questions, the possibility

that the most reliable performers of right action, in virtue of the traits of

character that qualify them as such, do not directly apply the Principle of

Utility in deciding how to act in the circumstances they face. This possibility

should be evident, for it should be evident that a person who develops such

traits may come, as a result, to pursue and live by goals and values whose

realization requires his following principles other than the Principle of

Utility. Fromthis observation it follows that the theory�s answers toquestions

about right actionsmay not always correspond to the decisions about what it

is right to do that the most reliable performers of right actions make. And if

the answers that do not correspond to their decisions include the embarrass-

ing prescriptions the theory yields, then themost reliable performers of right

actions would neither affirm nor act on those prescriptions. The prescrip-

tions, in otherwords,would havenouptake in their lives. Theywould neither

be regarded by such people as sound advice nor correspond to judgments

about what to do that such people reached when deciding what to do. Thus,

insofar as implementing the theory involved giving people the right motives

fromwhich to act and so making them the most reliable performers of right

action they could be, it is possible that the embarrassing prescriptions the

theory yields represent merely theoretical truths. At the final step of the

argument, therefore, one observes that the embarrassing prescriptions

might have no practical import within the theory and that they would, in

that event, cease to be a serious problem for it.


Of course, to turn this argument into a successful defense of act utilita-

rianism, defenders of the theory must do more than point to the possibility

that the most reliable performers of right action would apply other princi-

ples than the Principle of Utility in deciding how to act. Theymust also show

that this possibility is a strong one. They must show, in other words, that

there is strong reason to believe that the most reliable performers of right

action would directly apply other principles than the Principle of Utility in

deciding how to act and, in particular, that they would apply the basic

standards of justice and honesty in making such decisions. Otherwise

their defense would fall short of rendering plausible its crucial idea that

the most reliable performers of right actions would not follow principles

that yielded embarrassing prescriptions for the theory. It would fall short,

that is, of proving what it must prove to nullify the practical import of those

prescriptions. And failure to achieve this result would then mean that the

prescriptions continued to represent a serious problem for the theory. It is

reasonable to suppose, however, that defenders of act utilitarianism can

achieve this result. For there does appear to be strong reason to believe that

themost reliable performers of right actionwould follow principles that did

not yield any of these prescriptions.

Mill�s points about the indispensability of secondary principles to carry-

ing out the Principle of Utility suggest why. Recall, in particular, his point

that because of inescapable limits on human powers of foresight, discern-

ment, and calculation, it is necessary for people to use secondary principles

to achieve the end that the Principle of Utility prescribes. Even the most

powerful human minds, Mill would say, are liable to some misperception,

misjudgment, and miscalculation when applying the Principle of Utility

directly to circumstances in which it is not obvious what the best course

of action is. They, like those of us with less powerful minds, are therefore

more likely to determine the best course of action in such circumstances by

using generalizations reflecting the collective wisdom of people who have

found themselves in similar circumstances. Consider, as an apt illustration,

the difference between the way a great chess player like Gary Kasparov

determines his moves and how Deep Blue, the powerful chess-playing

computer that once defeated him in match play, determines its moves.

Kasparov, owing to the limits on his powers of foresight, judgment, and

calculation, necessarily makes his decisions by using good strategic rules of

chess, the game�s secondary principles, as it were, to analyze complex

Utilitarianism 117

positions and determine the best moves in them. Deep Blue, by contrast,

because it is not similarly limited in its powers of computation, can com-

pute an enormous number of possible outcomes for each of themoves it can

make in a given complex position and can also retain, without loss, its

memory of these possibilities. Consequently, it can analyze such positions

far more extensively than can Kasparov, and accordingly it can determine

the best moves tomake by direct calculation of whichmove has the highest

probability of having outcomes that will lead to victory. Plainly, if the most

reliable performers of right action, as defined by the Principle of Utility,

were like Deep Blue, they, too, would directly calculate which course of

action is likely to produce the most good. They would not, then, use secon-

dary principles to determine the best action. But because they are, like

Kasparov, human thinkers, there is strong reason to believe, instead, that

they would apply secondary principles to make such determinations.

Further, it is reasonable to believe, in view of Mill�s other point about the

indispensability of secondary principles to the utilitarian system of ethics,

that the secondary principles they would apply would include the basic

standards of justice and honesty. Mill�s other point, you will recall, is that

the enjoyment of certain essentials of human well-being like peace, secur-

ity, and liberty requires that all those who live together in some society

generally accept and comply with the basic standards of justice and hon-

esty, so it stands to reason that following such standards is necessary to

bringing about the most good overall for humankind. Hence, it stands to

reason that themost reliable performers of right action would include such

standards among the secondary principles they apply in determining what

it is right to do. Surely it would be odd, given the importance of the general

acceptance of and compliance with such standards to the good of human-

kind, to discover that the most reliable performers of right action applied

different standards.

8. Is act utilitarianism self-refuting?

Let us grant, then, that defenders of act utilitarianism could succeed in

saving the theory from the embarrassing prescriptions it yields if their

initial argument for seeing how these prescriptions might lack practical

import within the theorywere cogent. Accordingly, we are granting that the

most reliable performers of right action, as defined by the Principle of


Utility, would directly apply other principles in determining what it was

right to do in the circumstances they faced and that these other principles

would include the basic standards of justice and honesty. These concessions

allow us to narrow the question of the success of these defenders� efforts to

save act utilitarianism to that of whether their initial argument, the argu-

ment that raises the possibility that the embarrassing prescriptions the

theory yields would have no practical import, is cogent. Could it truly be

said of these prescriptions that they lacked practical import within the

theory despite their being consequences of the Principle of Utility? Or

would such a statement be false or even incoherent?

The statement would clearly be false, of course, if it were assumed that

the most reliable performers of right action believed the Principle of Utility

to be the fundamental principle of morality. For if they believed this, they

would then regard themselves as following other principles only because

doing so made them more reliable adherents to the Principle of Utility and

not because these other principles were basic determinants of right action.

They would therefore understand these principles, the basic standards of

justice and honesty, in particular, as strategic rules and not as rules that

placed limits on their conduct. Consequently, they could not ignore a

prescription to act dishonestly when they found themselves in situations

in which they were confident that they could bring aboutmore overall good

by dishonest action than honest action. They could no more ignore such a

prescription than Kasparov, analyzing a position in the middle of some

game, could ignore a conclusion he reached by direct calculation of the

position�s possibilities that a particularmove available to himwould lead to

certain victory. Even if he saw that the move broke some strategic rule of

chess, this would not dissuade him from making the move, for strategic

rules are not binding. And if he were confident of his analysis and reasoning

in this case, then it is themove hewouldmake. By the same token, themost

reliable performers of right action would not be dissuaded from doing

actions that, by their own lights, were certain to bring about the most

good overall, even if they knew that such actions would violate basic stand-

ards of justice and honesty. For the standards, when understood as strategic

rules, are not binding. If, having found a lost purse containing a hugewad of

cash, they were confident of their calculation that donating the cash to a

worthy charity would bring about more good overall than returning it to its

owner, then they would treat the prescription to donate the cash as

Utilitarianism 119

dispositive. Hence, the embarrassing prescriptions the theory yields would,

at least in some cases, have practical import for them.

So our sophisticated defenders of act utilitarianismmust assume that the

most reliable performers of right action do not believe the Principle of

Utility to be the fundamental principle of morality. They must assume,

then, that the most reliable performers of right action regard other princi-

ples than the Principle of Utility as the basic determinants of right action.

Asked what determines whether an action is right or wrong, a reliable

performer of right action will answer by citing these other principles,

including, in particular, the basic standards of justice and honesty. And he

will, at the same time, deny that what makes an act right is its bringing

about as much good for humankind overall as any other act that he could

perform in the circumstances. Hence, the theory, on this defense of it,

requires the advocacy and teaching of a different ethical theory for the

purpose of developing in people the traits of character that will make

them the most reliable performers of right action they can be. For people

must be kept ignorant of what truly makes some actions right and others

wrong if they are to develop these traits. The theory therefore requires, as a

condition of its implementation, that people be kept frombelieving that the

Principle of Utility is the fundamental principle ofmorality. On this defense

of act utilitarianism, the theory calls for its own suppression.

This is a paradoxical result, to be sure. Indeed some critics of act utilita-

rianism once thought it showed that the theory was self-refuting. But the

theory would be self-refuting only if morality was essentially public, and it

would be hard to argue against act utilitarianism on the basis of this thesis

without begging the question. For although many philosophers hold con-

ceptions of morality on which its principles are analogous to the laws of a

society and so are essentially public, defenders of act utilitarianism do not.

Rather they maintain a conception of morality on which the question of

whether to make a principle of morality public – whether to advocate or to

teach it – is a question about what action to takewith regard tomaking such

principles public, and like all such questions, it is answerable on the utili-

tarian theory according to the consequences of the alternatives open to one.

Thus, whether to make the Principle of Utility public is a question to which

one can apply the principle itself, and the answer, needless to say, depends

on whether publicity or obscurity best promotes the general good. That

suppressing rather than publicizing the principle could turn out to be the


action that best promotes the general good and therefore that the principle

might require its own suppression may seem paradoxical. But the theory

explains away the paradox when it makes clear that one can treat the

principle like any truth and ask whether propagating knowledge of it is

more conducive to the general good than keeping knowledge of it secret.

Hence, the defense of the theory we are presently considering can with-

stand this criticism.

9. When act utilitarianism ceases to be an ethical theory

The defense contains a deeper problem, however. Its initial argument, the

argument on which we are focusing, severs the very connection between

right action and reasonable action that it is the business of ethical theories

to explain. The connection is straightforward. Every ethical theory presup-

poses some notion of a competent moral agent, a person whose values and

judgment are sound, and of such agents it assumes that when they act

rightly, they act reasonably and conversely. A successful explanation of

the connection is thus a vindication of right action as reasonable action.

Were it otherwise, were the theory to employ a notion of a competentmoral

agent that failed to meet this condition, then the theory in answering

questions of how such agents ought to act would sometimes be caught

between giving answers advising that they act unreasonably in order to

act rightly and giving answers advising that they act wrongly lest they act

unreasonably. The theorywould then have failed to vindicate right action as

reasonable action. It would have failed as a practical discipline. When

sophisticated defenders of act utilitarianism suppose that the theory takes

the most reliable performers of right action as competent moral agents,

they suppose, as a result, that the theory allows, as a possibility, that those

people whose values and judgment are sound are people who do not follow

the Principle of Utility as the fundamental standard of right and wrong

action but follow other principles instead. Consequently, it is possible that

the theory, on their defense of it,must sometimes in answering questions of

how such people ought to act advise either that they act wrongly or that

they act unreasonably. For it is possible that sometimes the most reliable

performers of right action can act rightly only if they act contrary to their

principles, and it would not be reasonable for someone to go against his

principles if the values those principles reflected were sound. Thus in

Utilitarianism 121

supposing that the theory takes themost reliable performers of right action

as competent moral agents, our sophisticated defenders of act utilitarian-

ism sever the connection between right action and reasonable action, and

in doing so they render the theory incapable of vindicating right action as

reasonable action.

In effect, this defense of act utilitarianism treats the theory as one that is

not the product of a practical discipline. That it does so is clear from its

having severed the connection between right action and reasonable action,

for severance of the connection means that its being right to do some act

does not imply that morally competent agents have good and sufficient

reason to do the act. The upshot of the defense, then, is that it puts act

utilitarianism out of the business of ethical theory and into a different line

of work.


5 The moral law

1. Two theories of moral law

Teleological conceptions of morality originated in ancient Greek philoso-

phy. The major systems of ethics among the ancient Greeks, those of Plato

and Aristotle, in particular, were teleological. So too were those of Epicurus

and other thinkers who founded important schools of philosophy in the

period that came after Plato and Aristotle. Deontological conceptions, by

contrast, have a different origin. They derive from an ideal of universal

divine law that Christianity drew from the Judaic materials from which it

sprang. Christianity, to be sure, drew from the ancient Greeks as well. Its

identification of universal divine laws with the laws of nature, for instance,

comes from the Stoics, chiefly through Cicero (106–43 BC). But the ideas in

Christianity that yielded deontological conceptions are found in its under-

standing of divine laws as the laws of a supreme ruler that bind his subjects

to obey him in the way that a covenant with him would bind them. These

juristic ideas, which originated in Mosaic law, are the original frame for

deontological conceptions. The principal text that inspired them is Paul�s

statement in Romans: �When Gentiles who have not the law do by nature

what the law requires, they are a law to themselves even though they do not

have the law. They show that what the law requires is written on their

hearts, while their conscience also bears witness and their conflicting

thoughts accuse and perhaps excuse them.�1

The core thought of this passage is that humanbeings, simply by virtue of

being human, have an innate capacity to know right from wrong and to be

moved through their consciences, the seat of that knowledge, to do what is

right and to forbear doing what is wrong. This knowledge is the knowledge

of the requirements of law, and these requirements are written �on [our]

1 Romans 2.14–15 (Revised Standard Version).

123

hearts,� as Paul says. Their being so written means that we can have knowl-

edge of them through reflection on what is in our hearts. For this reason

none of us needs to be familiar with any holy book to have this knowledge.

Exercising one�s rational and reflective powers is sufficient. There is, there-

fore, a distinction to which Paul alludes, between knowing the law through

scripture and knowing it through reason and reflection. The former is

knowledge through revelation, and the latter is knowledge through reason.

Ironically, then, this central tenet of Christian thought makes recourse to

the Bible or any other religious text unnecessary for having knowledge of

right and wrong.

Knowledge of right and wrong is thus, according to core Christian belief,

knowledge of the requirements of moral law. But how can one know the

requirements of such law without recourse to a text in which that law is

written? How can one use reason apart from consulting such a text to

discover this law? One answer is through the study of nature itself and,

more specifically, human nature. If the laws in question are laws of nature,

then the study of how they regulate human life, its growth and develop-

ment, its maintenance in a healthy and thriving state, and its eventual

decline and expiration will yield an understanding of how one ought to

live. On this answer, one achieves the highest good by living according to

nature. In Christian thought, the laws of nature are God�s directives for

realizing this good, for nature, being the work of God, is the realization of

his goodness. That goodness guarantees harmony in how all the parts of

nature work together. And since the laws of nature are what give harmony

to the universe God has created, it stands to reason that men and women,

being part of this creation, achieve goodness in their lives by living as nature

(which is to say, God) intended. This is to say that men and women achieve

goodness by guiding their conduct by the laws that manifest these inten-

tions. This answer, however, plainly entails a teleological conception of

ethics and not a deontological one. Indeed, the conception it entails was

the dominant conception of the Middle Ages. To arrive at a deontological

conception we must consider a different answer, one that emerged in the

seventeenth century among a more lawyerly group of Christian thinkers.

This second answer, the deontologists� answer, derives from God�s being

the supreme ruler of the universe aswell as its creator. To conceive of God as

the supreme ruler of the universe is to understand him as sovereign over

humankind. That is, being God�s subjects, human beings are required to


obey him. This obedience not only is required by God�s sovereignty but also

makes sense, given the natural conditions of human life. For humans, while

naturally sociable creatures, nonetheless have difficulties in those condi-

tions sustaining society. To do so we need rules that we recognize and obey

in common, yet we would not recognize and obey such rules in common if

we did not mutually accept them as having authority over the conduct of

our lives. This mutual acceptance of the rules� authority follows immedi-

ately when God is understood to have issued them. His sovereignty over

humankind entails it.

The difficulties that human beings in their natural state have in sustain-

ing human societies are due to the conflicts that inevitably arise among

people who are largely concerned with promoting their own interests and

those of their loved ones. This is not to say that the natural conditions of

human life are so harsh and the self-seeking of men and women so single-

minded as to pit each of them against every other in a struggle for survival.

One need not, that is, suppose such harshness and self-seeking as Hobbes

supposed. Such a supposition was important to Hobbes�s ethics because it

supplied the basis onwhich he could derivemoral laws from a fundamental

principle of egoism. But an understanding of moral laws as rules God issues

directly to men so that they can live together peaceably, because it does not

rest on egoism, does not require suppositions like those Hobbesmade about

the natural conditions of human life. It would be enough to suppose that the

conditions produced quarrels and created frictions among people thatmade

their living together in society precarious. Rules restraining people from

engaging in conduct that provokes or aggravates quarrels and rules requir-

ing conduct that reduces friction, if they command sufficient obedience,

make collective life more peaceable and less precarious. Hence, from an

understanding of God as supreme ruler of the universe and of human beings

as sociable creatures who nonetheless need some order imposed on their

lives to live together peaceably, one can directly infer the idea of God�s

having given men and women laws general obedience to which establishes

that order.

Implicit in this idea is the thought of there being in human society a

moral order that is independent of the society�s positive laws and customs.

Every human society, according to this thought, instantiates this order,

even if the society�s positive laws and customs are contrary to it. Indeed,

what determines the justness and rightness of those laws and customs is the

The moral law 125

degree to which they conform to themoral order that God�s laws constitute.

Appeals by Christian social reformers to higher law to justify their disobe-

dience to positive law – the appeals to higher law of nineteenth-century

American abolitionists, for example, to justify their violation of laws regu-

lating slaveholding – reflect this thought. It is possible, then, to conceive of

human society as existing prior to positive law and social custom, prior to

government and culture. This conception goes hand in glove with an

inquiry into the standards of right and wrong that looks to the natural

conditions of human life for evidence of those standards. The hypothesis

of a human society existing prior to government and culture leaves those

conditions in place as the sole factors in determining the rules that create

order in a society. Hence, one can discover what the moral law requires by

reasoning from this hypothesis. In other words, one can have knowledge of

the moral law by reflecting on what rules human beings, in the absence of

government or culture, would need to obey in common to live peaceably

together in society. The second answer thus follows from this hypothesis.

The critical difference between the two answers lies, then, in how each

conceives of the relation between knowledge of what the moral law

requires and knowledge of the actions one must do to achieve the highest

good. The first answer, the answer of the scholastics, asserts that there is an

intimate relation between the two types of knowledge. On the scholastics�

answer,moral laws are laws of nature, and as such they give harmony to the

universe that God created. They give harmony both to the universe as a

whole and to the functioning of each thing that inhabits it. This harmony is

the laws� distinctive contribution to the goodness of God�s creation.

Something therefore works well when it works in accordance with the

laws that regulate the functioning of things of its kind. The moral law

regulates the functioning of human personality, since human psychology

is a phenomenon of nature. Consequently knowledge of the moral law is

knowledge of how to live well as a human being. It is knowledge of how to

achieve the highest good for human beings.

By contrast, the second answer, the answer of the deontologists, asserts

that the two types of knowledge are separate. On the deontologists� answer,

God imposes moral laws on men and women. The imposition is necessary

because the natural desires and emotions of human beings are divided

between the self-regarding and the social, and the conflicts among people

due to the former prevent the agreement among themdue to the latter from


being sufficient to sustain a society. In short, human beings, while social

creatures, are not like bees. They are only partly driven by their social

instincts, and these instincts do not dominate the self-regarding ones. To

the contrary, if anything, the reverse is true. As a result, external restraints

on the self-regarding desires and emotions of human beings are necessary

for sustaining a society. Moral laws supply those restraints. They are in this

sense external to nature. Obedience to them is required because of God�s

sovereignty and not because of their pointing the way to achieve the ulti-

mate end of human action. Indeed, sometimes obedience necessitates cur-

tailing one�s pursuit of the ultimate ends that are implied by one�s

self-regarding desires. Knowledge of the moral laws, therefore, on the

deontologists� answer, is a separate matter from knowledge of the highest

good. The one does not depend, not even implicitly, on the other.

The ideal contained in the deontologists� answer is that of one�s living in

fellowship with others as one�s equals under the rule of law. This ideal

corresponds to the notion we considered earlier of a moral order that is

independent of a society�s positive law and customs, its government and

culture. The rule of law itself is of course an ideal. It is the ideal of a

community whose members, to the last man or woman, are subject to its

laws. No member stands above them or is exempt from their authority. But

the ideal implicit in the deontologists� answer goes further. It also includes

the ideal of equality under the law: that all men and women are equally

subjects of the law and equally protected by it. In the case at hand, this

means that all are equally subject to and protected by God�s rule. The

relation between him, as their sovereign, and them, as his subjects, is the

same no matter who the subject is. In particular, no subject holds a higher

rank than any other in his or her relation toGod. None enjoys privileges that

others lack. It follows, then, that all distinctions of rank among human

beings, like those of nobility and commoner, aristocrat and plebeian, lord

and vassal, have no basis in themoral order that God�s law constitutes. They

are, in all instances, artifacts of the positive laws and customs in the

societies in which those ranks are recognized. And the same is true of

regimes of male supremacy, white supremacy, or the supremacy of one

ethnic group over another that have arisen throughout history in different

civilizations. All such regimes are contrary to the norms of equality that are

contained in this ideal. Thus the ideal that the deontologists� answer con-

tains is one that implies a community in which everyone bears the same

The moral law 127

basic duties towards one another and towards the community as a whole

and in which even the humblest member is accorded the same respect as a

duty-bearing member of the community as any of its more powerful

members.

2. Divine command theory

The deontologists� answer is based on a version of divine command theory.

In its most general form, divine command theory is the theory that identi-

fies right actions with actions that God has commanded and wrong actions

with actions that God has forbidden. The principal idea is that compliance

with God�s will is what makes an act right and noncompliance with his will

is what makes it wrong, and God�s commands express his will. In the

version of the theory on which the deontologists� answer is based, those

commands are understood as laws that God, as the supreme ruler of the

universe, addresses to humankind. This understanding corresponds to the

modern conception of laws as the commands that a sovereign addresses to

his subjects. Hobbes�s definition of �law� in Leviathan illustrates this concep-

tion. �[L]aw in general,� Hobbes writes, �is not Counsel but Command; nor a

Command of any man to any man; but only of him whose Command is

addressed to one formerly obliged to obey him.�2 AndHobbes completes the

conception by connecting the idea of a command with that of the will. Thus

his definition of �command� is �where a man says, Do this, or Do not this,

without expecting other reason than the Will of him who says it.�3 God�s

laws, then, on this version of divine command theory, express his will, and

therefore, following the theory�s principal idea, his will determines univer-

sal standards of right and wrong.

Hobbes�s definition of law includes a distinction between commands and

counsels. A counsel, Hobbes writes, �is where a man says, Do, or Do not this,

and deduces his reasons from the benefit that arrives by it to him to whom

he says it.�4 The point of the distinction is to sharpen the understanding of

law that Hobbes�s definition conveys. For in defining laws as commands as

distinct from counsels, Hobbes makes crystal clear that when a ruler issues

a law to his subjects, he means for them to obey it because he has com-

manded them andnot because they can deduce fromhis issuing the law that

2 Hobbes, Leviathan, ch. 26, par. 2. 3 Ibid., ch. 25, par. 2. 4 Ibid., ch. 25, par. 3.


he thinks theywill benefit from acting as he has commanded. Laws, in other

words, are authoritative because they express the ruler�s will and not

because they represent wisdom the ruler has about actions that his subjects

will benefit from doing. The point is evident in the case of bad laws. A bad

law, a law, for example, that a corrupt ruler of an impoverished country

enacts to enrich himself at the expense of his subjects is still authoritative in

that country as long as the ruler legitimately holds power and has enacted

the law in accordance with the protocols for legislation in that country�s

system of government.

That such laws are authoritative in that country does not mean of course

that anyonewho is subject to the authority of the country�s ruler acts rightly

in obeying those laws and wrongly if he disobeyed them. He acts legally in

obeying themandwould be acting illegally if he disobeyed them, butwhat it

is legal to do in a country is not necessarily what it is right to do, and

similarly, an action that it would be illegal to do in a country is not neces-

sarily an action that it would be wrong to do. Huck Finn, for example, acted

illegally in helping Jim in his attempt to escape slavery, but he did not act

wrongly. A country�s laws, therefore, though they are authoritative for all

who are subject to the authority of the country�s ruler or rulers, are not for

those same subjects standards of right and wrong. The possibility of bad

laws enacted by corrupt rulers makes this point clear.

But the same point, you might think, could not possibly apply to God�s

laws. In Christian thought, in particular, the authoritativeness of God�s laws

necessarily coincides with their being universal standards of right and

wrong. To assume otherwise would be to assume that God could make bad

laws, and because of God�s attributes, specifically, his being a loving God

who is all-knowing and all-powerful, that is not possible. God could not be

corrupt; he could not misjudge what laws would best serve the interests of

humankind; nothing could prevent him from imposing such laws on his

subjects. Being a loving God, he has the good of all human beings as his end

when he acts to impose laws on them, and being infinitely wise and

all-knowing his judgment as to what is best for human beings is error-

proof. Consequently, the laws he imposes are the best laws for establishing

a moral order within human society. There can be no basis, therefore, to

doubt that they are the standards of right and wrong for all human beings.

Nonetheless, the point does reveal a problem in divine command theory.

The theory�s principal idea is that what makes an act right is its being in

The moral law 129

compliance with God�s will and what makes it wrong is its opposing his will.

But if it is necessary to invokeGod�s being loving and all-knowing to show the

impossibility of God�s imposing bad laws on humankind, then the reason

that actions that comply with God�s will are right and actions that oppose his

will are wrong is not that God has commanded the former and forbidden the

latter. Rather it is that one can deduce from his having commanded the

former and forbidden the latter that he judges that the former are actions

human beings will benefit from doing and the latter are actions human

beings will suffer harm as a result of doing, and that his judgment in either

case could not possibly be wrong. In other words, if the theory must invoke

God�s being loving and all-knowing to certify the coincidence of the author-

itativeness of his commandswith their being universal standards of right and

wrong, then God�s telling human beings to do this and not to do that – e.g., to

love their neighbors as they love themselves and not to kill –may aswell be a

matter of his giving counsels as commands. Compliance with God�s will in

that case is not what makes an act right; opposition to his will is not what

makes it wrong. For the act would still be right even if God had only coun-

seled men and women to do it, wrong even if he had only counseled against

their doing it. Hence, to invoke God�s being loving and all-knowing to show

the impossibility of his imposing bad laws onhumankindmeans that one has

abandoned the principal idea of divine command theory.

Unfortunately, abandoning the theory�s principal idea means giving up

the thesis that God, by virtue of the commands he issues to humankind, is

the author of universal standards of right and wrong. It is to give up the

thesis that standards of right and wrong originate in God�s will. The reason

why is plain. To identify God�s commands with universal standards of right

andwrong because they are the commands of a loving and all-knowing ruler

is to suppose that God, in issuing those commands, takes the good of

humankind as his end and imposes laws according to a standard of effec-

tiveness, through organized human action, in realizing that end. It is thus to

suppose that God, in determining which acts to command and which to

forbid, follows a standard of right action that exists independently of his

will. But to acknowledge such a standard is to acknowledge a foundation of

morality that is independent of God�s will. All other standards of right and

wrong would then derive from this foundational standard and likewise be

independent of God�s will. These standards, while they coincide with God�s

commands, do not originate in his will.


Adivine command theoristmust thereforemaintain the theory�s principal

idea if he is to hold on to the thesis that morality springs fromGod�s will. Yet

maintaining this idea means holding that God�s will precedes every standard

of right and wrong. It means, that is, holding that no standard of right and

wrong exists independently of God�s will or serves to guide the judgments he

makes in issuing commands. And because God is all-powerful, there are no

limits on what he can command. These propositions constitute the doctrine

of theological voluntarism to which divine command theory, in its most

clear-eyed formulation, reduces. It follows then that no matter what act

God commanded, it would be right to do that act and wrong not to do it.

For no command of God could fail to be authoritative for human beings, and

therefore no command of God could fail to be a universal standard of right

action. Such is the import of theological voluntarism. Consequently, however

horrific or bizarre the act, if God commanded it, then it would be right to do it

and wrong not to do it. If God commanded, say, that blue-eyed infants be

tortured to death, then it would be right to torture blue-eyed infants to death

and wrong not to. But this result throws the theory into a real muddle. No

one, after all, would allow that someone would be acting rightly in torturing

infants to death, yet of course Christian belief in God requires taking God�s

commands as standards of right and wrong. The only way out of the muddle,

therefore, is to give up theological voluntarism. It is to give up the thesis that

the standards of right and wrong originate in God�s will.

3. Rational intuitionism

Giving up theological voluntarism does not, to be sure, mean that Christian

thought must give up its conception of God as the supreme ruler of the

universe. What it means is that, to retain this conception, a Christian

thinker must suppose that God imposes laws in accordance with standards

of right and wrong that exist independently of his will. That is, a Christian

thinker must suppose that such standards are evident to intelligent beings

and therefore to God, whose intelligence is infinitely greater than human

intelligence, and that God�s rule over human beings accords with his knowl-

edge and understanding of these standards. At the same, it is understood

that their validity is independent of his will and that his actions can be

judged by them. What is more, because God cannot err in determining his

will, the standards necessarily inform it and therefore the commands that

The moral law 131

he issues to human beings. These are the laws he imposes on humans, and

indeed he is often praised for the justice of his laws. Such praise, it is worth

noting, entails judgment of his action and therefore presupposes standards

of justice that exist independently of his will. For praising God�s rule over

human beings as just would be an empty encomium if there were no stand-

ards of justice independent of that rule by which to judge it. As Leibniz once

observed, why praise God for acting in accordance with justice if calling his

actions �just� says nothing more than that he acts?

The upshot, then, of this conception of God, the conception of him as the

supreme ruler of the universe whose laws are just, is that ethics does not rest

on premisses about God�s will. Indeed, it need not rest on any theological

premisses. The point was controversially made by Hugo Grotius (1583–1645),

when he declared that human beings would have the duties to one another

that God�s law imposes, �even if we should concede that which cannot be

conceded without the utmost wickedness, that there is no God or that the

affairs of men are of no concern to him.�5 Grotius, in this famous passage

from his great work, On the Law of War and Peace, at once affirmed God�s

position as supreme ruler of the universe and abandoned theological volun-

tarism. The two points combined to yield a view of ethics as grounded in

propositions or principles whose truth or validity, while present to God�s

intellect, is independent of his will.

It will be useful to compare the intellectualism to which Grotius� decla-

ration leads with the view we considered earlier that identifies the stand-

ards of right and wrong with the laws that a loving God, who is all-knowing

and all-powerful, imposes on human beings. This earlier view is most

naturally formulated as a theological version of rule utilitarianism. As

such, it includes theological premisses. In particular, it includes a premiss

about God�s love of humankind. It identifies the highest good with the well-

being of the object of God�s love and attributes to God thewill to bring about

the realization of this good to the greatest extent possible. Thus, one can

obtain a version of rule utilitarianism once one uses the theological premiss

about God�s love to identify the well-being of humankind as the highest

good, since from this identification and God�s being sovereign over human-

ity one can infer that the laws God imposes on human beings are such that

5 Hugo Grotius, On the Law of War and Peace, F.W. Kelsey, trans. (Oxford: Oxford University

Press, 1925), p. 13.


general compliance with them best realizes that good. In imposing these

laws on human beings, God acts in accordance with a principle of efficiency

or optimality in the realization of the highest good. And as the view in

question opposes theological voluntarism, the principle is understood to

exist independently of God�s will. In this regard, the view shares with the

intellectualist view of ethics that Grotius advanced a commitment to prin-

ciples of practical reasonwhose validity does not depend onGod�s existence

or his being concerned with the affairs of human beings. At the same time,

this version of rule utilitarianism departs from Grotian intellectualism in

taking God�s love to be what determines the good whose realization God

brings about through the imposition of laws on human beings. For in doing

so it represents God�s existence and concernwith human affairs as essential

to there being universal standards of right and wrong.

Of course, a version of rule utilitarianism that is fully consistent with

Grotian intellectualism is also possible. On this version, God imposes laws

on human beings in accordance with the Principle of Utility, which is taken

as the fundamental standard of morality and also as existing independently

of God�s will. That is, the Principle is understood to be evident to God as the

fundamental standard of morality but not to depend for its validity on his

existence or his concern with the affairs of human beings. Still, while this

version of rule utilitarianism is possible, it is not the most credible theory

within the intellectualist program that Grotius� view inspired. The reason is

that our best gauge of whether a standard of right, when taken as a funda-

mental standard, is evident to God is that it is evident, on reflection, to us,

and the Principle of Utility is decidedly less evident to us, on reflection, than

other standards of right and wrong. The difficulties with it that we saw in

the last chapter are ample proof of the uncertainty it brings. In this respect,

the instability we noted in rule utilitarianism is particularly telling. The

instability, recall, consists in a tension between the deontological character

of the theory�s secondary principles and the logic of utilitarianism, and this

tension is due to our having greater confidence in the validity of the

secondary principles than in that of the Principle of Utility. Consequently,

because the main thrust of the intellectualist program that Grotius� view

inspired is to fix, as the fundamental standards of morality, those that, on

reflection, seem self-evident, the most credible theories that have devel-

oped within the program are those identifying as fundamental those stand-

ards that agree more with common sense than does the Principle of Utility.

The moral law 133

In particular, it is a matter of common sense that acts are sometimes

right in themselves regardless of their consequences. Nothing opposes

utilitarianism more, for instance, than the common-sense maxim fiat justi-

tia, ruat coelum. Similarly, common-sense morality condemns lying in the

harshest of terms and seldom makes exceptions on account of its good

effects. The most credible theories, then, within the Grotian program iden-

tify a plurality of standards as fundamental to morality, many of which

correspond to features of an action that can make it right apart from its

having good consequences. Such features include an action�s being a piece

of speech that expresses the truth, its being the timely returning of an

object that one had borrowed, its being a showing of gratitude for advan-

tages another has freely given one, and its being the performance of an

action one had promised to do. The theories, then, identify as fundamental

standards of right and wrong, not only standards of nonaggression and

charity toward others and a standard of care for one�s own well-being, but

also standards of justice, honesty, fidelity, and gratitude. How different

standards are identified and enumerated varies from theory to theory.

Some theories identify a small number of very general standards each of

which comprehends several more specific right-making features. Others

identify a larger number and, accordingly, take each more specific feature

as corresponding to a separate fundamental standard. These differences are

unimportant, however. What is important is that all theories in this pro-

gram identify a plurality of fundamental standards of right andwrong some

of which qualify an action as right apart from its good consequences.

The program�smain epistemological supposition is that these fundamen-

tal standards, being objects of God�s intellect rather than deliverances of his

will, can be grasped and affirmed through the exercise of one�s rational

powers. That is, one can see, from reflection on what rules people must

observe in their relations to each other, the validity of certain standards,

and seeing the validity of these standards, one is compelled to affirm them.

What is more, on this supposition, one sees their validity immediately and

does not infer it from considerations of the means necessary to achieving

peaceful social relations. Accordingly, their validity is self-evident, and as

such one grasps it by exercising the intuitive powers of reason rather than

its inferential powers. The theories that developed from this supposition are

thus themselves products of a general epistemological theory aptly called

rational intuitionism. On this theory, the moral order that the fundamental


standards of right and wrong constitute is an object of the intellect, and

one�s knowledge of it comes from directly perceiving its elements and

structure through the exercise of the intuitive powers of reason.

The theory�s debt to Plato�s theory of forms is obvious. In either, the

objects of the intellect are held to be distinct from the objects of sensory

experience and known without recourse to the latter. Similarly, just as on

Plato�s theory one�s knowledge of the forms consists in grasping ideas that

are necessarily true, so on rational intuitionism one�s knowledge of the

moral order consists in recognizing standards of right and wrong that are

necessarily valid. Such knowledge, in either case, is different from empiri-

cal knowledge, which concerns contingent facts. It is different, that is, from

our knowledge that Mercury is the planet in the solar system whose orbit is

closest to the sun. That fact is contingent on the events occurring eons ago

that brought about the formation of the solar system. If those events had

been different, theremight now be a planet whose orbit is even closer to the

sun than Mercury�s. By contrast, knowledge of the forms, on Plato�s theory,

and knowledge of themoral order, according to rational intuitionists, is not

knowledge of things that could have been different. The intuitionists�

rejection of theological voluntarism makes this clear. Even if God had

commanded the torture of all blue-eyed infants, such torture would still

have been wrong. As the intuitionists might put it, its being wrong is not

contingent on anything, not even God�s will.

4. Ethics and mathematics

The force of rational intuitionism lies in the great confidencewe have in the

truth of such simple matters of right and wrong as that torturing infants is

wrong. The same confidence is at work in our immediately seeing serious

problems with act utilitarianism in its yielding prescriptions that conflict

with basic standards of justice and honesty. Our verdicts in either case seem

unshakeable. Of course, we are confident, too, in our belief that Mercury is

the planet in the solar system whose orbit is closest to the sun. But that

confidence would vanish if astronomers from the world�s leading observ-

atories and research universities were to announce the discovery of a

planet whose orbit lies entirely within Mercury�s. By contrast, no similar

announcement about the newly discovered permissibility of torturing

infants, even if it were made by leaders of the world�s major religions and

The moral law 135

past winners of the Nobel peace prize, would shake our confidence in the

truth of its being wrong to torture infants. Our confidence in this truth is

not merely greater, then. It is, so it seems, of a different order.

Rational intuitionism explains this difference by sharply distinguishing

between the objects of intellect and the objects of sensory experience and

attributing necessity – one might say metaphysical necessity – to the truths

about the former and contingency, metaphysical contingency, to the truths

about the latter. Yet the problem with this explanation is that the attribu-

tion of necessity to truths about matters of right and wrongmay be nothing

more than a reiteration of our great confidence in their being true. And

similarly for the intuitionists� companion thesis that we have knowledge of

these matters through the exercise of intuitive powers that enable us to

apprehend directly their truth, that their truth is self-evident to reason.

Either thesis, in its own way, expresses a belief in the certainty of one�s

judgment on simple matters of right and wrong, but it offers nothing

further to back up that certainty.

What is needed, then, is support for these theses that does not depend

on our confidence in the truth of certain propositions of right and wrong

and our recognizing that this confidence comes from something other

than confidence in the methods of empirical science. To this end rational

intuitionists have adduced mathematics as evidence of the type of truths

and knowledge that they hold moral truths and moral knowledge to be.

Surely, they argue, the necessity of such mathematical propositions as

the simple truths of arithmetic, like 2 + 5 = 7, will be granted. And surely

it will be granted, too, that such truths are known without recourse to

sensory experience. Further, in so describing them, we are not merely

expressing belief in the certainty of our mathematical judgments or

reiterating our confidence in their being true. Rather our description

captures an understanding of mathematical objects that anyone familiar

with the methods of the discipline, methods that are as old as Euclid, will

have acquired. Anyone, that is, who is familiar with these methods will

have come to understand mathematical objects as objects of pure intel-

lect whose existence and nature could not be other than they are.

Accordingly, mathematics furnishes us with the right model for under-

standing the necessary existence and character of moral standards and

the certainty of moral knowledge. Or so rational intuitionists have often

maintained.


The principal method that they have invoked is that of developing a

branch of mathematics as an axiom system. Such a system consists of a

set of propositions a small number of which are the axioms or postulates,

the fundamental propositions of the system, and the rest of which are

theorems, propositions that one can derive from the postulates and previ-

ously derived theorems. The paradigm of the method is Euclidean geome-

try, and for centuries thinkers followed Euclid in taking the postulates of

geometry to be self-evident truths and the procedures by which one proved

the theorems of geometry to be truth preserving. The self-evidence of the

postulates and the truth-preservingness of the proof procedures meant that

the theorems of Euclidean geometry were truths whose necessity was as

certain as the necessity of the postulates. And one could be certain of the

postulates because, being self-evident, one could directly apprehend their

truth through exercising one�s intuitive powers of reason. The match

between this common understanding of geometrical truth and knowledge,

on the one hand, and the understanding of moral truth and knowledge that

rational intuitionism promotes, on the other, should be plain. For rational

intuitionism takes the fundamental standards of right and wrong to be like

the postulates of Euclidean geometry, treats other moral principles as

analogous to theorems of geometry, and sees the moral order these stand-

ards constitute as an object of the intellect on a par with the geometrical

space that the Euclidean postulates define.

Unfortunately for rational intuitionism, advances in mathematics begin-

ning in the latter half of the nineteenth century overturned this under-

standing of axiomatic treatments of a branch of mathematics. In particular,

mathematicians no longer regard the postulates of an axiomatic system as

self-evident. They gave up this view once they began to construct and

explore alternative systems within a single branch of the discipline. The

most influential of these constructions are the axiomatic systems for geom-

etry that are alternatives to Euclidean geometry. In each of these alterna-

tives, the crucial difference between it and Euclidean geometry is found in

its postulate about parallel lines. In Euclidean geometry the relevant postu-

late is Euclid�s famous parallel postulate, that given a point p and a line l that

does not contain p, there exists one and only one line that contains p and is

parallel to l. In a non-Euclidean geometry this postulate is replaced by a

postulate that contradicts it either by denying that there is any line parallel

to l that contains p or by holding that there is more than one. Since a

The moral law 137

proposition and any proposition that contradicts it cannot both be true,

they cannot both be self-evident. Hence, entertaining an axiomatic system

that is an alternative to Euclidean geometrymeans giving up the idea that to

be a postulate a proposition must be self-evident. Of course, one could still

insist that Euclid�s parallel postulate is self-evident and that the contra-

dictory propositions that are postulates in non-Euclidean systems are

mere logical possibilities, as are the systems themselves. But such insistence

would amount to no more than a mere expression of one�s belief in the

certainty of Euclid�s parallel postulate, a reiteration of one�s confidence in

its being true, and would not represent anything mathematics supports.

The same points, moreover, hold true of axiomatic treatments of the

discipline�s other branches.

The consequence of these advances inmathematics is devastating for the

appeal that rational intuitionists make to mathematics to support their

theses about the necessity and self-evidence of fundamental standards of

right and wrong. In view of these advances, as we�ve seen, these intuition-

ists cannot offer the postulates of an axiomatic system in mathematics as

evidence of propositions whose truth one can, through exercising the

intuitive powers of reason, directly apprehend. Nor can they adduce the

axiomatic method as evidence of the necessity of mathematical truth.

The reason is that the method upholds only the necessity of the theorems

of an axiomatic system relative to the necessity of the postulates from

which those theorems are derived. It does not uphold the necessity of the

postulates themselves. To be sure, this point need not stop rational intui-

tionists from continuing to attribute necessity to these postulates if they

like. But should they continue, it would be rather unclear what this attri-

bution means in the case of postulates whose contradictories are logically

possible. Calling it mathematical necessity would be vacuous. Calling it

metaphysical necessity would be unilluminating.

There remains the sharp distinction rational intuitionists draw between

objects of the intellect and objects of sensory experience. That distinction,

to be sure, is unaffected by the construction of alternative axiomatic sys-

tems within a single branch of mathematics, since one can still distinguish

the methods of mathematics from the methods of empirical science and

take the subject of the former to be objects of pure intellect. At the same

time, we can now see that an important assumption about these objects

that is implicit in the rational intuitionists� appeal to mathematics is


unwarranted. Because the intuitionists explicitly assume that the postu-

lates of a branch of mathematics are necessary truths, they must further

assume, albeit implicitly, that the objects defined by those postulates are

the only objects with which that branch deals. Obviously, the construction

of alternative axiomatic systems within a single branch makes this implicit

assumption untenable. No one nowadays would even suggest that geome-

try, for instance, deals exclusively with geometric objects in Euclidean

space and that the distinctive postulates and theorems of non-Euclidean

geometries are simply false propositions about the properties of those

objects. The point by itself merely reinforces our earlier criticism of rational

intuitionism�s appeal to mathematics. But it should also remind us that

rational intuitionists must be making an analogous assumption about the

objects of ethics and that this assumption, too, is unwarranted. Its being

unwarranted, moreover, spells even deeper trouble for the theory.

Specifically, rational intuitionists must be assuming that the objects of

intellect that the fundamental standards of right and wrong define –

particularly, the moral order that those standards constitute – are the

only objects with which ethics deals. Yet nothing warrants this assump-

tion. This point may at first seem odd. If the fundamental standards of

right and wrong are universal, if every human society, regardless of its

laws and customs, instantiates the moral order they constitute, then that

order and its related objects must be the subject of ethics. So how could

this assumption be unwarranted? But the question rests on a confusion of

two kinds of study, a confusion that revisiting the analogy with geometry

should clear up. If we understand geometry as a study of certain objects

of pure intellect, then it is a study of geometric objects and the different

forms of space, Euclidean and non-Euclidean, that contain them. These

different forms of space correspond to alternative axiomatic systems.

Apart from this study, there is also the study of physical space, for the

universe exists in space and one can ask what the form of the space in

which the universe exists is. Specifically, one can ask, �Which of the

different spaces, Euclidean and non-Euclidean, does the universe

instantiate?� The study that this question initiates is not a study of objects

of pure intellect. It is rather a study that requires recourse to our sensory

experience of the position and motion of material objects. By analogy,

then, rational intuitionists must distinguish within ethics the study of

objects of pure intellect, specifically, the different forms of moral order

The moral law 139

that could be instantiated in human society, and the study of the form

that is so instantiated. The latter study, given that the fundamental

standards of right and wrong are universal, must be a study of the one

moral order that every human society instantiates. But nothing warrants

the assumption that the former study deals with only one set of universal

standards and thus with only one form of moral order. And it is this

assumption, which rational intuitionists, by virtue of their taking the objects

of ethics to be objects of pure intellect, make, that is unwarranted.

They need the assumption, however. They need, that is, to avoid the

possibility of there being alternative sets of fundamental moral standards

and thus different forms of moral order. For this possibility would open the

door to asking what determined which of these forms was the moral order

that every human society instantiates, which of these alternative sets of

standards was the set of fundamental standards of right and wrong that we

recognize and follow. In the context of Christian thought, it would open the

door to asking what explains God�s choice of the set of standards in accord-

ancewithwhich he issues laws to human beings. Clearly, the answer cannot

be that God chooses to issue laws in accordance with the set of standards of

right andwrongwhose validity is self-evident, for as objects of pure intellect

no set stands apart from the others as self-evident, no more than the

postulates of one axiomatic system within a branch of mathematics stand

apart from the postulates of the alternative systems as self-evident. It

appears, then, that the distinction between objects of intellect and objects

of sensory experience provides no support for the rational intuitionists�

theses about the necessity and self-evidence of the fundamental standards

of right and wrong. Without some support for these theses, however, or

warrant for the assumption that there is only one form of moral order with

which ethics deals, rational intuitionism is in danger of collapse. And in the

context of Christian thought such a collapse would amount to a reversion to

theological voluntarism.

5. Kant�s way

Immanuel Kant (1724–1804) shared the rational intuitionists� belief that

knowledge of right andwrong was knowledge of the requirements of moral

law and that such knowledge was a matter of common sense. On Kant�s

and the intuitionists� view, moral knowledge does not require a deep


understanding of human nature or complicated calculations of the conse-

quences of the various actions open to one in the situations one faces. Of

course, Kant and the intuitionists granted that sometimeswe face situations

in which it is hard to know what the right thing to do is. But for the most

part, they held, we have no trouble determining in the situations we face

what the moral law requires. Our ordinary powers of reason, they main-

tained, are sufficient for this purpose.

Kant broke with rational intuitionism, however, on the question of how

we know, through the exercise of reason, what the moral law requires.

According to rational intuitionism, the moral law corresponds to funda-

mental standards of right and wrong, and these standards are self-evident.

We directly see their validity through use of our intuitive powers of reason.

That is, we know directly and by intuition that we ought not to kill human

beings, that we ought to tell the truth, that we ought to keep our promises,

and so forth, and we then apply these precepts to the situations we face.

Kant, by contrast, did not appeal to the intuitive powers of reason to explain

our knowledge of the moral law. He did not think we directly saw the

validity of fundamental standards of right and wrong that corresponded

to that law. Rather he thought that each of us knows whether an act is

required by the moral law, whether we have a duty to do it, by a process of

reasoning about it. The process is one of practical thought as distinct from

speculative thought, a distinction that is of the utmost importance in Kant�s

philosophy and of no importance to the rational intuitionists�. Kant under-

stood the process of reasoning by which we determine whether an act is

required by moral law as embedded within other processes of practical

thought. Accordingly, he drew his account of our knowledge of right and

wrong from a general account of the operations of practical reason. And it

will therefore be necessary to set out this general account of practical

reason before examining Kant�s account of our knowledge of right and

wrong.

In Kant�s view, the workings of practical reason in human action are

what set the actions of human beings apart from all other actions and

events in nature. Human beings, Kant held, are rational agents. Reason,

that is, has a practical as well as theoretical function. Its powers include the

power to direct action. Thus, according to Kant, when aman acts rationally,

he acts in accordance with a rule or principle that, through the exercise of

practical reason, he sees as applying to his situation and suited to his ends.

The moral law 141

�Everything in nature,� Kant wrote, �works in accordance with laws. Only a

rational being has the power to act in accordancewith his idea of laws – that

is, in accordance with principles – and only so has he a will. Since reason is

required in order to derive actions from laws, the will is nothing but

practical reason.�6 This passage from Kant�s Groundwork of the Metaphysics of

Morals begins his celebrated argument for understanding the moral law as

the promulgation of reason.

What does Kant mean by acting in accordance with principles? In the

first instance and as a first approximation, he means acting according to a

generalization that connects an action in a given situation with a result that

the agent intends. The result is therefore the agent�s end, and the general-

ization captures the agent�s understanding of the action as, in the situation

he is in, a goodmeans to this end. Aman, for example, who turns on a faucet

with the intention of washing his hands acts on the generalization that

turning the handle of a working faucet releases water from the supply to

which it is attached. His end is to get water for washing hands, and his

means is turning a handle on a faucet attached to a supply of water large

enough for that purpose. Getting water, needless to say, is only his imme-

diate end. He also has the end of washing his hands, preventing infection or

perhaps the ingestion of disease-producing microbes, and ultimately being

happy. All of these ends are objects of human desire, and while different

people have different desires, that is, different tastes and interests, hopes

and ambitions, each person desires his or her own happiness as an ultimate

end. It is, in that individual�s life, the ultimate end consisting in the sat-

isfaction of his or her individual tastes and interests, hopes and ambitions.

Practical reason, then, in the first instance, guides human beings in the

pursuit of happiness by providing, in the form of principles, intelligence

about the best means to achieving their ends.

Kant introduces the term �imperative� to denote certain principles of

practical reason. It is a technical term in his philosophy. It applies to those

principles with which reason dictates compliance. The dictatorial character

of imperatives is essential to Kant�s understanding of them. Because human

beings have appetites and passions, because many of our desires and

6 Immanuel Kant, Groundwork of the Metaphysics of Morals, H. J. Paton, trans. (New York:

Harper & Row, 1964), p. 412. References are to page numbers in the Preussische

Akademie edition.


emotions originate in animal instinct, we often act on those desires and

emotions without giving enough thought to the consequences and implica-

tions of our actions. Such is particularly likely when a person is subject to

strong desires or in the grip of a powerful emotion. In these cases, the

person acts on poor judgment but not necessarily against reason.

Sometimes, though, a person does not act merely on poor judgment but

irrationally. That is, the person knows that reason requires his taking a

certain course of action, yet he fails to take it. You know, for instance, that

to get to your chemistry class on time, youwill have to leave your apartment

by 9:45, and there is no question of your wanting to be in class on timemore

than your wanting to remain in your apartment reading your favorite blog.

Yet 9:45 comes and goes, and you continue to read the blog. In this case,

your leaving at 9:45 is a requirement of reason. As Kant would see it, reason

does not merely propose a 9:45 departure as the best course of action, but

demands it as the only course, given that your end is to be in class on time.

Departing at that time is necessary for achieving this end. Such necessity is

what qualifies the principle you would be acting on, if you acted rationally

in this situation, as an imperative.

Kant distinguishes different types of imperative according to the kind of

necessity they imply. The type in our example is what Kant calls a hypo-

thetical imperative. It is hypothetical because the necessity of taking the

action is conditional on the agent�s having the end relative to which the

action is a means. In the example above, given that you want to be in class

on time, it is necessary that you leave your apartment by 9:45. But it

wouldn�t be necessary if you had a different end whose achievement did

not also require a 9:45 departure. The imperative, in other words, that you

leave your apartment by 9:45 is valid only if the end relative to which your

leaving by that time is a means is an end that you in fact have. Let us call

such an end a validating end. Accordingly, Kant distinguishes, within the

class of hypothetical imperatives, between two further types of imperative,

technical and pragmatic, according as the validating end is the end of an

activity or enterprise that human beings engage in at their discretion or is

an end of an activity or enterprise that all human beings, just in virtue of

their having natural desires, are engaged in. An example of the former

would be the imperative, with regard to cooking a hard-boiled egg, that

one cook the egg in boiling water for more than three minutes. An example

of the latter would be any hypothetical imperative whose validating end

The moral law 143

was the agent�s happiness, for in Kant�s view all human beings pursue

happiness, not because they choose to but because it is part of their nature

as human beings. Consequently, the necessity with which pragmatic imper-

atives dictate action is a higher grade than the necessity with which tech-

nical imperatives dictate action.

Neither, however, dictates actionwith the highest grade of necessity. The

highest grade would be the necessity of an imperative whose validity was

not conditional on any of the agent�s ends. The highest grade, that is, would

be an imperative whose validity was unconditional. Kant calls such an

imperative a categorical imperative. It is distinct from every type of hypo-

thetical imperative. His bold thesis is that moral laws are categorical

imperatives.

This would not be such a bold thesis, of course, if all that Kant meant in

identifying moral laws with categorical imperatives was that moral laws

were the commands of God, for there would be nothing bold in character-

izing God�s commands as unconditionally valid. Hobbes�s distinction

between commands and counsels, to which we referred earlier, makes

this clear. While the validity of a counsel depends on the ends of those to

whom counsel is given, the validity of a command does not. It depends

instead on the commander�s having authority over those to whom the

command is given, and God, being the supreme ruler of the universe, has

unconditional authority over human beings. Accordingly, his commands

are unconditionally valid. But Kant meant something else. Given his special

use of the term �imperative� to mean principle of practical reason that

dictates action, he meant that the moral laws were principles of practical

reason and valid as such rather than as expressions of an authority�s will.

Hence, being categorical imperatives, they are unconditionally valid as

principles of practical reason rather than as expressions of a supreme

ruler�s will. Consequently, any rational agent, regardless of his or her

ends, would be acting contrary to reason if he or she violated a moral law.

And this is a bold claim.

How could one substantiate such a claim? Or as Kant put the question in

the Groundwork, how is it even possible that there are categorical impera-

tives?7 It is evident, Kant pointed out, how there can be hypothetical imper-

atives. Someone who decides to pursue a certain end thereby sets himself to

7 Ibid., p. 419.


take whatever means are necessary to achieving that end. This person may

later, to be sure, on realizing the lengths to which he would have to go to

achieve this end, decide that achieving it isn�t worth the trouble, but at that

point he will have also abandoned his decision to pursue it. In other words,

as long as it is his will to pursue this end, it must be his will to takewhatever

means are necessary to achieving it. There would be an incoherence in his

will if this were not so. Hence, an imperative that directed him to take this

means makes sense as a requirement of reason. But how could there be any

incoherence in his thinking or his will if he decided to act dishonestly or

cruelly in the pursuit of some end? By what process of reasoning could a

person be led to see the necessity, as a matter of coherent thinking or

willing, of acting honestly or forbearing from acting cruelly? Kant answers

these questions by taking such actions to be lawless and arguing that only

by keeping one�s actions lawful can one avoid incoherence either in one�s

thinking or in one�s will. Categorical imperatives are possible, then, accord-

ing to Kant, because lawfulness itself can be a condition of rational action.

Kant�s idea, in proposing lawfulness as a condition of rational action, is

that the agent, in acting on principles, act only on those principles that

could be made into laws for all human beings. That is, we are to test the

principles on which we act to see whether they could be adopted as univer-

sal laws, and if they can, then our acting on them is lawful action. Moreover,

Kant understands this test as purely formal. A principle of action does not

fail the test just because it is predicated on the pursuit of a certain, sub-

stantive end. Rather, failure occurs only when there is something incoher-

ent either in the conception of the principle as a universal law or in the act

of will required tomake it a universal law. To Kant�s way of thinking, as long

as you are able to form the idea of everyone�s acting on the principle of your

action when they find themselves in situations similar to yours and as long

as you can will consistently with your decision to act on this principle that

you live in a world in which everyone acts on the principle in situations like

yours, then your acting on the principle is lawful. When you cannot form

the idea because of some incoherence in it or, though you can form it, you

cannot will its realization consistently with your decision to act on the

principle, then acting on the principle is unlawful. In either case, the

incoherence or inconsistency makes it necessary that you forbear from

acting on the principle. The imperative, then, that you not act on it is

categorical. Or rather it would be categorical if acting lawfully in Kant�s

The moral law 145

sense, that is, acting on principles that could be made into laws for all

human beings, were a condition of acting rationally.

Kant, as I said, proposed that it was. He put his proposal in the form of a

second-order principle. It is second order in that it is a principle regulating

the first-order principles on which rational agents act. The first-order prin-

ciples are the generalizations that connect an action in a given situation

with an intended result and that an agent follows in pursuing an end to

whose advancement the intended result contributes. Kant calls these first-

order principles maxims.8 Accordingly, the second-order principle he for-

mulates is this: Act only on that maxim through which you can at the same time will

that it should become a universal law.9 This principle, given that it represents a

requirement of practical reason, is itself, then, an imperative that dictates

with unconditional necessity. Indeed, Kant in recognition of its being the

source of all particular categorical imperatives or imperatives of duty refers

to it singularly as the Categorical Imperative. It is thus, in his system, the

fundamental principle of morality. Fidelity to it in our deliberations about

what acts it would be right to do, he believes, leads us to make moral

judgments coincident with common sense. Faithful application of the

Categorical Imperative in one�s deliberations, that is, not only shows what

the moral law requires but also reinforces the judgments of ordinary men

and women about right and wrong.

Kant�s identification of moral laws with categorical imperatives turns,

then, on whether lawfulness is a condition of rational action. Kant did

not think it would be easy to show that it is or that if he did, he would

have established as a matter of general psychological fact that human

beings, when they act honestly, honorably, kindly, and the like, act on

categorical imperatives rather than on other motives. He did maintain

that anyone who firmly believed in the validity of universal standards of

right and wrong and did not think they were, as he put it, �chimerical�

must accept his conception of them. But firm belief falls short of knowl-

edge, and it is knowledge that Kant sought. Still, we can at this point see

whether his conception of universal standards of morality as categorical

imperatives is cogent and leave for later, if it is, examining the complex

case he made for his proposal that lawfulness is a condition of rational

action.

8 Ibid., p. 421n. 9 Ibid., p. 421.


6. Formalism in ethics

The main objection to Kant�s conception is that it is too formal to be a useful

guide to telling right action from wrong action. �Empty formalism� is the

standard charge that its critics have raised against it. The Categorical

Imperative directs rational agents to determine whether they can will that

the maxims of their actions become universal laws, and the procedure for

determining this is purely formal. It involves no substantive criterion of right

and wrong. On this procedure, the agent first identifies the maxim of his

action, whether it be an action he has already done or one that he is contem-

plating doing. The agent then determineswhether he can forma coherent idea

of a world in which everyone whose circumstances are like his does the same

action.That is,hedetermineswhetherhecan formacoherent ideaof aworld in

whichhismaximhas become auniversal law regulating the conduct of human

beings in such circumstances. And finally, supposing he can form this idea, he

determines whether he canwill that such aworld be created consistently with

hiswilling the action. The criteria the agent uses to test themaxim�s suitability

for being willed as a universal law are coherence and consistency. These are

formal criteria, and such criteria, Kant�s critics havemaintained, cannot alone

yield definite conclusions about whether an action is right or wrong. While

Kant, to be sure, meant the procedure to be one that yielded definite conclu-

sions about the rightness or wrongness of a course of action one had taken or

wascontemplating taking,onecan in fact, so thecriticismcontinues, always rig

things at the procedure�s first step so that themaximone identifies at this step

passes the tests that one applies at the following steps. Hence, Kant�s critics

conclude, the procedure for applying the Categorical Imperative – let us call it

the CI procedure – is ultimately arbitrary.

Let us take up this charge of arbitrariness first. It will be useful, in seeing

how serious the charge is, to work with a concrete example. Our example of

your finding a lost purse containing a huge wad of cash will do nicely. Recall

that youfind thepurse, notice the cash, and recognize that you ought to return

both to its owner. It would be wrong, you conclude, to keep the cash for

yourself. ButhowcantheCIprocedureexplainyour coming to this conclusion?

Its first step requires that you identify themaxim youwould be acting on

if you kept the cash. Let us assume that if you did keep the cash, your reason

would be to make yourself richer. Accordingly, the maxim of your action

would be something like the following:

The moral law 147

(M1) Given circumstances in which I can appropriate for myself

something valuable that belongs to another without fear of punishment or

retaliation, I will take the valuable thing in order to enrich myself.

At the next step, you form the idea of a world in which everyone who finds

himself in a situation in which he can, without fear of punishment or

retaliation, enrich himself by appropriating something valuable that

belongs to another does so. This idea, you should then immediately realize,

is incoherent. It is incoherent because a world in which no one respects the

property of others would be a world in which, in effect, there would be no

property and none of the rights of ownership that go with it, and in such a

world people cannot appropriate for themselves things that belong to

others. The incoherence of the idea therefore tells you that you cannot act

on this maxim. It tells you, in other words, that it would be wrong to keep

the cash in order to enrich yourself.

Admittedly, what conclusion the procedure yields depends on the direc-

tive you identify as the maxim of your action. Suppose, for example, you

identified the following as your maxim:

(M2) Given circumstances in which I can appropriate for myself cash con-

tained in a lost purse that I find in a public park, without fear of punishment

or retaliation, I will keep the cash in order to enrich myself.

Taking this as your maxim, you would then, at step two of the procedure,

form the idea of aworld inwhich everyonewhofinds a lost purse in a public

park keeps whatever cash the purse contains whenever they can do so

without fear of punishment or retaliation. You would thus form the idea

of a world in which universal disrespect of the property of others was

limited to cash contained in lost purses found in public parks. Since univer-

sal disrespect that was so limited would not amount to the wholesale

dissolution of property rights, you would not see anything incoherent in

this idea. Nor need you discover any opposition in your will to the creation

of such a world that would be inconsistent with your deciding to keep the

cash so as to make yourself richer. M2 therefore passes both the test of

ideational coherence at step two and the test of volitional consistency at

step three. In other words, if it were the maxim of your action, the CI

procedure would tell you that it would not be wrong to keep the cash.

Obviously, if M1 and M2 were equally plausible candidates for being the

maximof your action, then the CI procedurewould be open tomanipulation.


In that case, you could, by deciding to identify one directive rather than the

other as your maxim, rig the procedure so that it yielded the conclusion you

wanted. For this reason critics have long chargedKant�s procedurewithbeing

arbitrary. But the charge is overblown. Using the procedure to determine

whether an act is right or wrong is not a game. It presupposes a genuine

interest in knowing themoral status of your act, whether it is right or wrong,

and therefore presupposes sincerity in identifying the maxim you would be

acting on in doing it. Given that you are sincere in identifying this maxim,

there should be few if any cases inwhich you are genuinely torn between two

formulations of it on each of which the CI procedure yields a conclusion

opposed to the conclusion it yields on the other.

Further, you can tell whether you have properly identified your maxim

by determining whether you have included in its formulation only those

features of the action and its circumstances that wouldmake a difference to

your decision about whether to do it. Thus, with regard to M2, it should be

clear on several counts that this would not be the maxim on which you

would be acting if you kept the cash. First, that you found the purse in a

public park makes no difference to your decision whether to keep the cash

it contains. Had you found the purse in an alley instead, the decision would

still be the same. Second, that the cash is contained in a purse is likewise

irrelevant. Your decision would not be affected by its being contained in a

wallet, a briefcase, or even a gym bag. Third, and what is most important,

your decision is not about keeping or returning someone else�s cash as such.

It is about keeping or returning something valuable that belongs to another

and that you have come to possess. Leaving aside questions of safety from

punishment or retaliation, your decision would be the same even if what

the purse contained were not cash but rare stamps, an endorsed cashier�s

check, or a winning lottery ticket. M1, then, because it does not, as it were,

descend to the level of irrelevant specificity about the action and its circum-

stances that M2 does, is far more plausible as a candidate for being the

maxim of your action.

The question of whether there could be an equally plausible candidate

that was even less specific than M1 is trickier. Thus consider the

following:

(M3) Given circumstances in which I can gain possession of something

valuable without fear of punishment or retaliation, I will take the valuable

thing in order to add to the stock of valuable things that I possess.

The moral law 149

If this were the maxim of your action, then again you could form a coherent

idea of a world in which everyone took possession of valuable things when-

ever they could do so without fear of punishment or retaliation. In such a

world there would be no property or rights of ownership, to be sure, but

M3, unlike M1, does not presuppose the existence of property or ownership

rights and therefore your forming the idea of a world in which it became a

universal law would not be similarly incoherent. Indeed, the idea you would

form would correspond to Hobbes�s account of the natural condition of

humankind, and no one has ever thought there was something incoherent

inHobbes�s account. At the same time,whether you could alsowill that such a

world be created consistently with a decision to keep the cash is questionable

at best, since in such a world you are likely to possess a smaller stock of

valuable things and a decision to take the cash manifests a will to possess a

larger one. But even before considering whether M3 passes the tests of idea-

tional coherence and volitional consistency at steps two and three of the

procedure, there is the question of whether M3 could even be your maxim.

It could be, if you were either completely innocent of notions of private

property and knew nothing about the rights that owners of property have or

had rejected these notions and denied that anyone could privately own any-

thing or have the rights that private ownership entails. Either possibility is

extremely unlikely, of course, but neither can be ruled outwithout ruling out

other instances of ignorance or rejection of an aspect of the moral culture in

which an agent finds himself that are not so unlikely. People, after all, are

sometimes ignorant of the moral culture that partly defines the circumstan-

ces of their actions, or they are sometimes sufficiently opposed to it as to have

rejected it wholly or in part. Recent immigrants from faraway lands are often

ignorant of much of their new country�s moral culture. Committed anar-

chists reject government and the laws that governments enact. The maxims

of their actions, in either case, may therefore be directives that exclude

references to cultural or legal aspects of their circumstances that people

who are at home in the culture and accept the authority of government and

the laws it enacts would include in the directives that are themaxims of their

actions. In general, then, an agent�s maxim of action includes reference to all

and only those features of his circumstances that are relevant to his decision

to act, and for agents who are familiar with the moral culture that partly

defines their circumstances and have not rejected any aspect of it that is

relevant to their decisions, their maxims must include reference to those


aspects of it. M1, therefore, is the maxim on which you would be acting if

you kept the cash in order to enrich yourself, except in the extremely

unlikely event that you are innocent of the ideas of property and ownership

rights or have rejected them. In that event, M3 could be your maxim.

7. The problem with Kant�s formalism

The CI procedure can thus survive the charge of arbitrariness that its critics

commonly make against it. Sincere identification of your maxim gives you

no leeway to rig the procedure. There remains, then, the more general

charge that the procedure is excessively formal. While this charge is vaguer

than the charge of arbitrariness, we can nonetheless understand it as at

least criticizing the CI procedure for lacking a substantive criterion of right

action. And there does appear to be something to this charge, for the

CI procedure is not, as turns out, completely reliable. Sometimes it yields

false negatives. That is, sometimes it yields a conclusion about its being

wrong to act on a maxim of action that, as a matter of common sense, it is

not wrong to act on. Moreover, that it yields such conclusions is due to the

purely formal notion of law that Kant assumed. This deficiency is most

readily seen in examples concerning the application of the second of the

procedure�s two tests, the test of volitional consistency. How Kant meant

this test to apply to an agent�s maxim is itself somewhat obscure, however.

So let us first, to clarify it, work through one of his primary examples.

Kant, in this example, imagines a prosperous man coming upon others

who are in great distress and whom he could easily help. This man, Kant

writes, �[then] thinks, �What does it matter tome? Let every one be as happy

asHeavenwills or as he canmake himself. I won�t deprive himof anything; I

won�t even envy him; only I have nowish to contribute anything to his well-

being or to his support in distress!��10 Implicitly, then, the man�s maxim is:

(N1) Given circumstances in which I can easily help others who are in

distress, I will withhold assistance to them in order to avoid the inconven-

ience to my own pursuits that it would entail.

Kant observes that such a maxim would pass the first test of the CI

procedure, for one can form a coherent idea of a world in which no one

freely inconveniences himself to help others who are in distress. But he

10 Ibid., pp. 423.

The moral law 151

maintains that the maxim would fail the second. The reason, he argues, is

that occasions may arise in this man�s life in which he will need the help

and support of others, and he would be depriving himself of that help if he

were, through his will, to bring about a world in which no one ever helps

others who are in distress and whom they can easily help. Hence, there

would be an inconsistency in his will if he were to act on N1. And from this

result it follows that he has duty to help the people in distress he has

encountered.

Though the argument as it stands has gaps, filling them is not a problem.

To begin with, let us observe that N1, when it is one�s maxim, manifests a

will to advance one�s happiness. Kant�s argument, then, must show that,

given N1 as the agent�s maxim, the agent would be thwarting his will if he

brought about, through it, a world in which no one helps others in distress

when he or she can easily do so. Showing that the agent�s will would be so

thwarted shows the requisite inconsistency in the will, an inconsistency

that consists in the agent�s deciding to act so as to advance his happiness

and at the same time willing the creation of a world in which his prospects

for happiness are significantly diminished. And to show this inconsistency

requires the supposition that a person, no matter how prosperous he may

be, at least tacitly knows that theremay be occasions in his life when hewill

need the freely given help of others and that without such help his life

would become significantly worse. In effect, then, Kant must be supposing

that people, however prosperous, at least tacitly know that they may even-

tually be the victims of accidents, mishaps, or crimes, that they may find

themselves lost in unfamiliar places or stranded far from home, that they

may become targets of unprovoked vengeance, that they may lose their

homes to floods,fires, tornadoes, or other natural disasters, and so forth and

that in any of these circumstances they would look to others for help and

hope to receive it. For knowing this, the person will recognize that his

prospects for happiness are better in a world in which people have goodwill

toward each other than in a world in which they are indifferent to each

other�s well-being and accordingly, given a will to advance his happiness,

he would will the creation of the former and oppose the creation of the

latter. To will the creation of the latter, in other words, would be incon-

sistent with his will to advance his happiness. Once we add this supposition

and its implications to the argument, Kant�s conclusion that N1 fails the CI

procedure�s test of volitional consistency follows directly.


It is sometimes said, in criticism of Kant�s argument, that it is an

essentially prudential argument for a duty of to help others. Yet this

criticism misses the point of the second test. The point of the test is to

see whether one�s will, if fixed by a given maxim, would be thwarted were

that maxim to become, through one�s will, a universal law. When the

given maxim is prudential, as N1 is, then the reasoning by which the

agent determines whether his will would be thwarted were his maxim to

become, through it, a universal law is perforce prudential. But the test itself

is not restricted to prudential maxims, and if the given maxim were, for

instance, the maxim of an agent seeking to promote world peace, then the

agent�s reasoning in applying the test would be oriented toward the

achievement of world peace rather than prudentially oriented. Kant�s

argument therefore appears to consist of prudential considerations in

favor of a duty to help others only because the maxim that fixes the will,

N1, is prudential. What the agent actually determines, when he sees that

his will would be thwarted if N1 were to become, through it, a universal

law, is that he would not be acting lawfully if he acted on N1. This consid-

eration is what establishes that he has a duty to help others in the situation

he faces. He discovers an inconsistency in his will, and realizes that

the only way to resolve it, if he is to act in accordance with law, is to

abandon N1.

It is now possible to see how the CI procedure overreaches in determin-

ing whether courses of action are right or wrong. Specifically, the condi-

tion that one�s maxim be suitable for being willed as a universal law of

nature places too great a constraint on the courses of action a person can

pursue without acting wrongly. The trouble arises because many courses

of action will not, in fact, be pursued by everyone, and to let considera-

tions of what the world would be like if everyone pursued a course of

action determine the rightness of that course is, in some cases in which

relatively few people would in fact pursue it, to exclude as wrong courses

of action that are in fact permissible.

Here is an example. Suppose you live in an area where the incidence of

violent crime has risen high enough to cause you some concern about going

out at night. To protect yourself you decide to carry a handgun concealed on

your person. Doing so, let us further suppose, is permitted by the state and

the smaller units of government that are responsible for public safety. The

maxim of your action would then be something like:

The moral law 153

(N2) Given circumstances inwhich I am in some danger of being the victim of

violent crime, I will carry a concealed handgun in order to protect myself

from this danger.

Following the CI procedure, you test this maxim by first seeing whether you

can form the idea of a world in which everyone who is in some danger of

being the victim of violent crime carries a concealed handgun. It is certainly

a coherent idea. Next you determine whether you canwill that such a world

be created consistently with your decision to carry a concealed handgun for

protection. At this point you should see that you cannot will that such a

world be created consistently with your will�s having been fixed by N2, for

in bringing about such a world, through your will, you would be creating

even more dangerous circumstances for yourself, ones in which you are in

fact at greater risk of being the victim of violence, criminal and otherwise.

This is because aworld inwhich everyonewho is concerned about being the

victim of violent crime carries a concealed handgun is a world awash in

handguns, and such aworld is amuchmore dangerous place than aworld in

which only a relatively small number of people carry handguns.

Consequently, you see that you would be thwarting your will if you brought

about, through it, such a world and must therefore conclude that it would

be wrong to carry a handgun to protect yourself. Yet surely it is not wrong

for someone to carry a handgun for protection, as long as the state and other

units of government responsible for public safety permit it.

The reason the CI procedure runs into trouble when applied to examples

like this one is that the notion of law implicit in the procedure�s tests is

entirely formal. It is the notion of a supremely authoritative rule by which

every rational agent is governed, and it is entirely formal in that it does not

contain the idea of the rule�s having an end or purpose beyond that of

prescribing lawful conduct. In particular, it does not contain the idea of

the rule�s having the purpose of helping to sustain human society by either

restraining people from doing things that, if done too often, would threaten

the society�s stability or requiring people to do things that, if not done often

enough, would also threaten the society�s stability. Rather the application

of either test appears to express an interest on the agent�s part in acting

lawfully for the sake of acting lawfully and not for the sake of participating

in a scheme for bringing about social good through general obedience to

law. In this respect, the notion of law implicit in the procedure�s tests is


more like Hobbes�s notion than a notion that implies a social purpose to law

as such. It is more like Hobbes�s notion in that the agent is not supposed to

look for reason to obey beyond the fact that he is subject to the law, subject,

that is, to the authority of the ruler who imposes it. As a result, the CI

procedure invites criticism similar to the one that proved fatal to theolog-

ical voluntarism. It invites the criticism that the unlawfulness of an action is

not itself a reason against doing the action if the relevant notion of law is

entirely formal, for such a notion does not preclude innocuous or sensible

actions from being unlawful.

Trouble arises, in particular, then in the application of the second test,

because whether one can, consistently with a decision to act on a certain

maxim, will the creation of a world in which that maxim is a universal law

will sometimes appear to be an idle consideration if one thinks the lawful-

ness of an actionmatters to its being reasonable to do only if the social good

would not be jeopardized by letting people do it. It is not an idle consider-

ation, of course, if one can expect people generally to be similarlymotivated

in similar circumstances and to forbear from acting on the maxim from a

shared understanding that the actionmust be forborne for the sake of social

good. But it does appear to be idle if one knows that few others would be

similarly motivated in similar circumstances, too few at any rate, for one to

regret the collective effect of their acting on that maxim. For in that case

letting people do the action will not jeopardize the social good.11

The charge of excessive formalism against Kant�s CI procedure has sig-

nificant merit, then, when it is understood as an objection to the formal

notion of law implicit in the procedure. The notion is implicated in the false

negatives that the procedure yields. It is also, on account of the notion�s

being divorced from any end or purpose that law has as an instrument of

maintaining order and stability in human society, hard to see what interest

11 One might, in light of this observation, think that the trouble arises only because N2

does not accurately express your maxim, since it omits from the description of your

circumstances that few people carry handguns. But obviously if your maxim included

this feature of your circumstances – that is, if your maxim were: (N3) Given circum-

stances inwhich I am in somedanger of being the victimof violent crime and only a few

people are carrying handguns, I will carry a concealed handgun in order to protect

myself from this danger, then you would again find that it would be unlawful to act on

thismaxim. The reason is that the conception of themaxim as a universal law of nature

is incoherent; you cannot coherently conceive of a world in which everyone carries a

handgun in circumstances in which few people do.

The moral law 155

a rational agent could have in acting lawfully except that of doing so for the

sake of acting lawfully. Such an interest, however, seems no more reason-

able than the interest of a political subordinate in blind obedience to his

political superiors. It borders on idolatry toward law. In any case, the notion

is markedly different from the notion that originally inspired the deon-

tologists� answer to the question of how human beings could know the

requirements of moral law through reason and without consulting a text in

which those requirements are laid down. That notion corresponded to an

understanding of morality as based on substantive criteria of right and

wrong that arise from a need to impose order on human society. Kant�s

notion, by contrast, corresponds to a different understanding of morality,

one that springs from his idea that lawfulness is a condition of rational

action. The idea, as we noted at the close of section 5, is not easy to show,

and our subsequent criticism of the CI procedure has only added to the

difficulty. Kant thus carries the heavy burden of having to vindicate this

idea. Unless he does, the charge of excessive formalism commonly made

against his ethics will stick.


6 The ethics of self-determination

1. Kant�s step into metaphysics

What interest could rational agents have in acting lawfully if not the order,

stability, and other collective goods that law brings to society? Why should

it otherwise matter to them that their actions are lawful? It wouldmatter to

them, of course, if acting unlawfully made them liable to punishment. But

in that case their interest in acting lawfully would not come from seeing it

as a good thing. It would come, rather, from seeing it as the surest way to

avoid a bad thing, something they have an interest in escaping. Yet the

challenge to an ethics like Kant�s that represents lawfulness as the essence

of moral action is to explain what could interest rational agents in acting

lawfully regardless of how the law is enforced, regardless, that is, of

whether it is enforced by threats of punishment or incentives to obey. The

question then that confronts a defender of Kant�s ethics is why a rational

agent should regard an action�s being lawful as a condition of its being

reasonable to do. If he cannot give an answer to this question, the charge

of excessive formalism will stick.

Kant himself was fully aware of the importance of this question. He

understood that a person must realize some value through acting lawfully,

else making lawfulness a condition of the reasonability of an action would

be pointless. It would have no rational basis. This value,moreover, had to be

recognized by all who possessed reason by virtue of their possessing reason.

It could not be due to a desire, sentiment, feeling, or other contingent

affection, not even one common to all human beings. For if it were, then

the point of insuring that one�s action was lawful would be due to some-

thing in the agent�s psychology that, being contingent, he could lack, and

therefore there could be circumstances, if only hypothetically, in which the

agent�s forbearing from an act because it was unlawful would be pointless.

That is, whenhe forbore from it in those circumstances hewould not realize

157

the value of acting lawfully that normally gives such forbearance a point.

Consequently, only if the point of insuring that one�s action is lawful is due

to a value of which one could be cognizant regardless of the desires, senti-

ments, feelings, and other contingent affections to which one is liable,

could the lawfulness of an action be, for any agent, a necessary condition

of its being reasonable to do. Hence, Kant observed, for categorical imper-

atives to be possible there must be a value that rational agents realize

through obedience to them, and it must be a value that these agents could

recognize regardless of the desires, sentiments, feelings, and other contin-

gent affections to which they are liable.

With this observation, Kant begins to build a deeper theory of morality.

Or as he puts it, he takes his first step intometaphysics.1 Before taking this

step, he had, as we saw, divided imperatives into hypothetical and catego-

rical, used this division to distinguish different degrees of necessity with

which an imperative dictates action, and then identified the Categorical

Imperative – the second-order principle whose application to first-order

principles yields imperatives of the highest degree of necessity – as the

fundamental principle of morality. Kant�s aim, in expounding these ideas,

was to fix the type of experience in which moral law appears to rational

agents and to describe the character of the thinking by which they typi-

cally determine what themoral law requires. Our experience ofmoral law,

according to Kant, is thus the experience of being directed by uncondi-

tional dictates of reason, dictates we determine by seeing whether in

taking a certain course of action we would be allowing ourselves more

liberty in the pursuit of our ends than we would permit others to have in

the pursuit of theirs. It is, in short, the experience of being bound by a

universal rule that has the backing of reason, and Kant�s step into meta-

physics initiates a theory of what in the nature of human existence must

lie behind and make possible such an experience if the experience is

not illusory.

The first step in the construction of this theory, then, is Kant�s introduc-

tion of a value that a rational agent realizes through acting lawfully. Some

such value is needed to guarantee that acting lawfully always has a rational

basis. And for there to be such a basis, Kant holds, the value realized through

1 Kant, Groundwork of the Metaphysics of Morals, p. 426.


acting lawfullymust have its source outside of the agent�s ends. Hence, Kant

is led to the idea of two radically different kinds of value in human life. On

the one hand, human beings find value in the satisfaction of their natural

desires and personal interests, the determinants of their ends as sensuous

beings. Accordingly, such value is relative to these desires and interests and

varies among human beings to the degree that the corresponding desires

and interests likewise vary. On the other hand, because human beings are

rational agents and thus capable of acting lawfully, they are all cognizant of

the value that lawful action realizes. This value provides them with a

rational basis for such action, and because it is a value that rational agents

recognize by virtue of their being rational, Kant characterizes it as absolute.

Unlike values that are relative to natural desires and personal interests, it

does not vary among human beings. It is constant for all rational beings and

necessarily so by virtue of there being no variation in reason across rational

agents.

Kant�s idea of there being two radically different kinds of value in human

life corresponds to his view of human beings as having a dual nature. It

corresponds to a view of human beings as both sensuous beings, beings in

whom sensory experience composes, at least partly, their mental lives, and

rational beings, beings in whom the processes of reason compose, at least

partly, their mental lives. Humans, in Kant�s view, are unlike beasts who, if

sentient, are wholly sensuous beings. At the same time, humans do share a

sensuous nature with such beasts. But possessing reason, they also have a

nature that beasts lack. This duality in human nature is reflected, then, in

two kinds of activity in human life, the pursuit of happiness and the

exercise of reason for its own sake, and through each of these kinds of

activity human beings realize one or the other of the two kinds of value that

Kant�s deeper theory of morality assumes.

Thus Kant regards the pursuit of happiness as the pursuit of things that

satisfy one�s natural desires and personal interests, and these things have

value for one as constituents of that happiness. Their value corresponds to

the contribution they make to one�s happiness, and the greater the contri-

bution the more they are worth to one. Accordingly, Kant characterizes

their value as a kind of price, for having measured the value of each by its

contribution to one�s happiness, one would be willing to forgo it – that is,

pay the price of not having it – in return for attaining something that would

contribute even more to one�s happiness. By contrast, Kant regards the

The ethics of self-determination 159

exercise of reason for its own sake as a self-sufficient activity. One needs no

external reward to engage in it. The value of engaging in it inheres in the

activity itself. There is value, Kant holds, to being a rational agent, and this

value has no price. For its value does not consist in its contribution to one�s

happiness, and hence there is no commonmeasure by which to compare it

with those things that do. The value realized through rational activity is

independent of, and indeed, incommensurable with, the value of the things

that contribute to one�s happiness.

The absolute value of rational activity or of being a rational agent there-

fore provides the rational basis of acting lawfully. Kant arrives at this crucial

tenet of his theory by reasoning hypothetically. Let x, he supposes, be that

which has absolute value and whose absolute value provides the rational

basis for acting lawfully. There must be such an x if there is a point to acting

lawfully. And, further, x must ground the same set of imperatives as the CI

procedure yields, else it would not explain the experience of being bound by

a universal rule backed by reason. It would not, in other words, explain our

experience of moral law. To explain this experience, then, the value of

x must be such that one realizes it through obedience to moral law as

popularly understood, for moral law, as popularly understood, is what the

CI procedure is meant to capture. And only if this value inheres in rational

activity itself or in rational agency itself could it explain this experience.

Thus Kant comes to identify our rational nature as a possible ground of all

categorical imperatives. x, he declares, is humanity or, more generally,

rational agency. It is that which, as an end in itself, grounds the moral law.2

2. The formula of humanity

But what does Kant mean by describing human beings, and rational agents

generally, as existing as ends in themselves? And does he mean something

more than that the value of rational agency is absolute? These questions

arise because the phrase �end in itself� is rather odd. Indeed, it is peculiar to

Kant�s ethics. No other philosopher before Kant used it. Rather his prede-

cessors, in referring to an action�s end, meant either a goal of the action, or

2 Or as Kant puts it, having supposed that there is somethingwhose existence has absolute

value, �Now I say that man, and in general every rational being, exists as an end in

himself, and not merely as a means for arbitrary use by this or that will.� Ibid., p. 428.


an effect that its agent intended to bring about, or an objective that the

agent hoped to accomplish. Kant, by contrast, does not mean any of these

things in describing human beings as existing as ends in themselves. He

does not mean that human beings are goals to be pursued or effects one can

intend to bring about, or objectives one can hope to accomplish. Indeed, he

could not mean any of these things since, procreation aside, none is a

coherent idea.

What is evenmore important, though, Kant could not mean any of these

things, because he understands the function of humanity�s being an end in

itself (or what for Kant comes to the same thing, our rational nature�s being

an end in itself) as the same as the function of the Categorical Imperative.

Like the latter, it applies to one�smaxims and not to one�s actions directly. It

is not, therefore, an end that one sets for oneself and chooses means to

achieve. It is rather a value that, being inherent in reason, regulates the

rational activity of setting ends and choosingmeans to them. Specifically, it

is a value inherent in reason that, if upheld, keeps one from setting ends the

pursuit of which would denigrate one�s own humanity or the humanity of

others. And likewise it is a value inherent in reason that, if upheld, keeps

one from choosing means the use of which would denigrate one�s own

humanity or that of others. In short, it regulates the activity of practical

reason by guiding rational agents toward setting ends and choosing means

that comport with the agent�s value as an end in himself and the like value

of those whose lives are touched by his actions.

Accordingly, Kant draws fromhis notion of humanity as an end in itself an

alternative formulation of the Categorical Imperative: Act in such a way that you

always treat humanity, whether in your own person or the person of any other, never

simply as a means but always at the same time as an end.3 Following a common

practice in recent Kant scholarship, let us refer to this formulation as the

formula of humanity and to the earlier formulation as the formula of univer-

sal law.4 Kant means these to be different formulations of one and the same

principle. He does not regard them as different principles. They are, in his

view, alternative formulations of the fundamental principle of morality, the

Categorical Imperative. It follows, then, that on his view, each formulation

yields in every case the same result as the other. Thus, on his view, if you

apply the formula of humanity to the situation in which you find a lost purse

3 Ibid., p. 429. 4 See above, p. 146.


containing a hugewad of cash in the bushes of a neighborhood park, youwill

come to the same conclusion as to what you ought to do as you would reach

in applying the formula of universal law. And similarly if you apply the

formula of humanity to the situation in which you happen upon a person

in immediate need of help whom you can, thanks to your own prosperity,

easily help, youwill come to the same conclusion aboutwhat you ought to do

as you would reach in applying the formula of universal law.

Kant�s viewmay seem surprising, even untenable. How could a principle

predicated on the idea of universal law be identical to a principle predicated

on the idea of humanity or rational nature as an end in itself? Neither idea

implies the other. They arewholly independent of each other. Sowhy aren�t

the two formulations similarly independent? And if they are, then in what

sense can they be formulations of one and the same principle? In the

strictest sense, of course, they cannot be, for two principles cannot be

identical if the concepts that constitute each are different from those that

constitute the other. Presumably, Kant would agree, for he clearly does not

mean that the two formulations are synonymous. That is, he does not mean

that they are formulations of the same principle in the sense that to for-

mulate Euclid�s parallel postulate as a postulate about there being, with

respect to a line l in a plane p and a point r also in p but not lying in l, one

and only one line in p that both contains r and is parallel to l and to

formulate it as a postulate about there being with respect to l and r only

one line in p that contains r and contains no point intersecting l is to give

two formulations of the same postulate. So it must be in some looser sense

that he means that they are the same. The most likely possibility is that he

conceives of the two formulations as equivalent in that each yields the same

conclusion as the other when applied to the same practical situation. Each

then serves equally well as the fundamental principle of morality. Or in

other words, taking morality to be founded on a single principle, one can

take either to be that principle without it affecting one�s judgments about

where one�s duties lie.

Some recent defenders of Kant�s ethics reject even this weaker thesis

about the relation between the two formulations of the Categorical

Imperative. They deny that the two are equivalent. They hold, instead,

that the formula of humanity, despite Kant�s own view of the matter, is

superior to the formula of universal law and in some situations yields

results differing from the results one gets from applying the latter.


Because the CI procedure is not, as we saw, foolproof, because it produces

false negatives, one attraction of this view is that it offers a reconstruction

of Kant�s ethics that avoids the problem of excessive formalism that these

false negatives represent. For instance, one could very plausibly argue that

if you apply the formula of humanity to the situation in which you are in

danger of being the victim of a violent crime, you will not reach the

conclusion that it is impermissible to carry a concealed handgun for the

purpose of self-protection, since neither will you be using anyone as a

means when you take such action nor will you be failing to treat anyone

as an end. By contrast, when you apply the formula to the situation inwhich

you find a lost purse containing a huge wad of cash, you will conclude that

you may not keep the cash since to do so would be to ignore that the cash

belongs to another and to treat its owner as merely a supplier of funds for

your use. It would be to treat her, in other words, simply as a means.

Similarly, when you apply the formula to the situation inwhich you happen

upon someone in immediate need of help and are prosperous enough to be

able to help, youwill conclude that you ought to help since to deny someone

in such distress help when you can easily provide it would be to fail to treat

him as an end.

3. Is the formula of humanity an independent principle?

This modern view of Kant�s ethics conceives of the formula of humanity as

an independent principle for assessingmaxims of action. In other words, on

this view, it is understood to provide separate and distinct criteria for

assessing maxims of action from those that the formula of universal law

provides. This understanding plainly departs from Kant�s own understand-

ing of the relation between the two formulae. For it is clearly Kant�s inten-

tion, in introducing the formula of humanity, to be developing a theoretical

substructure to the formula of universal law. It is clear, that is, that he

believes that there is a structure to moral thought in which the formula of

humanity is at a deeper level than the formula of universal law and which

his step intometaphysics is meant to uncover. So to conceive of the formula

of humanity as an independent principle is to abandon Kant�s own concep-

tion of it as a formula that undergirds the formula of universal law. One

immediate worry, then, to which this departure from Kant�s thought gives

rise is that the criteria for assessing maxims of action that the formula


provides are, when detached from those of the formula of universal law, too

vague to yield definite conclusions in more than a few stock situations.

Their vagueness is due, on the one hand, to the obscurity of Kant�s notion

of an end in itself and, on the other, to its being largely indeterminate

whether in one�s actions one is treating someone simply as a means.

What, for example, explains why you treat a news vendor as an end and

not simply as a means when you buy a newspaper from him? How does this

purchase differ from your buying the same paper from a machine, since in

that case you are not treating the machine as an end and since in both cases

the transactionmay consist in littlemore than your giving up three quarters

to the vendor or the machine and picking up a paper?

Suppose, as some would argue, that the reason why you treat the vendor

as an end but not the machine is that your transaction with him involves

mutual consent. The vendor�s consent implies that your exchange with him

satisfies some end of his, and your understanding that it does so is sufficient

for denying that you treat him simply as a means. But nowwemust wonder

about examples, like those of the gigolo and the gold digger, in which

someone takes advantage of another�s love without reciprocation. Because

of the one-sidedness of the relationships in these examples, we describe

them as ones in which one party simply uses the other. Thus amanwho has

no feelings for a woman but nonetheless takes advantage of her love by

allowing her to pay the rent on his apartment, put him through dental

school, buy him expensive gifts, and so forth is simply using her, we would

say, and we would not withdraw this description even if we knew that she

was not blind to his lack of feelings for her. Her consent to the arrangement,

though it implies that it satisfies some end of hers and though he under-

stands this, would still not alter our perception of him as simply using her as

a means to his living a more comfortable life. But how, then, does the

vendor�s consent differ?

Nor is this conundrum the only problem with determining what counts

as using another simply as a means. When a photojournalist takes photo-

graphs of people involved in events that are making news, war refugees,

say, or protesters at a public demonstration, is he using his subjects simply

as a means to reporting the news? It might seem so, especially since photo-

journalists seldomask their subjects for consent and since some of the latter

undoubtedly do not wish to be photographed. Yet Kant surely did not mean

to condemn such actions (or their eighteenth-century counterparts) as


contrary to moral law. Similar uncertainties arise with consideration of

such actions as those of someonewho is writing an unauthorized biography

of a famous person for the purpose ofmakingmoney from sales of the work

or people who bet on sporting events like tennis matches for the sheer fun

of it. Perhaps these are not examples of using people as ameans in the sense

that Kantmeant, because in none of them is someone beingmanipulated by

another into doing something he would not otherwise do. But if one cannot

use someone as a means in Kant�s sense unless one manipulates that

person, then the typical actions of peeping Toms and malicious gossips

will not be contrary to law according to the formula of humanity. And

who would deny that a man who spies on his neighbor for the excitement

of watching a stranger undress was using that neighbor simply as a means

to gratifying his prurient interests?

There is, moreover, a further worry about the view of Kant�s ethics on

which the formula of humanity is conceived of as an independent principle.

This worry is that such a view erases the difference between Kant�s ethics

and the ethics of the rational intuitionists. For unlike the formula of uni-

versal law, which one applies through the CI procedure, the formula of

humanity taken by itself does not appear to be similarly embedded in a

process of practical reason. There is no distinctive form of practical reason

that it structures. Rather one applies it to themaxims of one�s actions as one

would apply any substantive principle, such as a principle that prescribes

honesty and forbids deception or one that prescribes loyalty and forbids

betrayal. In other words, the use of the formula to determine whether one

may act on a given maxim entails no form of reasoning other than that of

applying a general principle to a particular object or fact, and this form of

reasoning is not peculiar to practical thought.

Perhaps those who have offered reconstructions of Kant�s ethics on

which the formula of humanity is independent of and superior to the

formula of universal law would object to this point on the grounds that to

determine whether someone is being treated as an end and not simply as a

means, because it requires thinking about ends and means, must make use

of the form of practical thought that means-to-ends reasoning entails. But

applying the formula does not in fact require such reasoning, for Kant�s

notion of someone�s being an end, his notion of humanity as an end in itself,

does not, as we noted earlier, imply the ordinary notion of an end, and

determining whether you are using someone simply as a means is not a


matter of determining ameans to an end but rather of determiningwhether

there is anything more to your using someone as a means to an end than

that you so use him. The upshot, then, of these recent attempts to recon-

struct Kant�s ethics appears to be not so much an improvement on Kant�s

ethics as a reversion to rational intuitionism.

How then shall we understand Kant�s conception of the formula of

humanity as depending on the formula of universal law? The answer to

this question lies in how the formula of universal law enters into the way

you determine whether you would be, in acting on a certain maxim, treat-

ing humanity in yourself or another as an end. For to treat humanity, either

in yourself or another, as an end requires honoring the dictates of reason.

This requirement follows from Kant�s distinction between humanity and

animality, a distinction reflected in his view of human beings as having a

dual nature. Thus Kant identifies humanity with our rational nature and

animality with our sensuous nature. So to treat humanity as an end is to

treat the rational nature of human beings as such, and one treats man�s

rational nature as an end by honoring the dictates of reason. Plainly, then,

one treats humanity in your own person as an end by acting only on those

maxims that pass the two tests the CI procedure comprises, for to act on a

maxim that failed one of these tests would be to disregard a dictate of reason

in order to satisfy some natural desire or personal interest, and in doing so

one would be subordinating one�s humanity to one�s animality, which is to

say, failing to treat the former as an end. Similarly, one treats humanity in

the person of another as an end by actingwith respect to this person only on

maxims that he would regard, if he were reasonable, as acceptable maxims

to act on. Andwhat determineswhether a reasonable personwould regard a

maxim as acceptable is the CI procedure, for it represents the deliverances

of reason on the permissibility of acting on the maxims to which it is

applied, and a reasonable person is one whose judgments correspond to

these deliverances.

Kant, it is worth noting, expressly makes it a condition of your not

treating another simply as a means that the latter agree with your

maxim.5 At the same time, Kant clearly does not mean by agreement with

a maxim consent to it. Criminals, he observes, do not as a rule consent to

their punishment, yet it does not follow that a judge who imposes

5 Kant, Groundwork of the Metaphysics of Morals, pp. 429–30.


punishment on a criminal who does not wish to be punished treats him

simply as a means. The criminal does not consent, but having a rational

nature, he is able to understand when punishment is just and lawful. He

knowswhat justice and the law demand of thosewho commit crimes even if

he does not exercise that knowledge at the time of his sentencing. He thus

agrees with the maxim on which the judge acts in imposing punishment

even if he refuses to acknowledge the judgment that demonstrates the

agreement. Or as I put this point in the paragraph above, he would regard

themaxim on which the judge acted as an acceptable maxim to act on, if he

judged it reasonably. Hence, one can say that the judge treats him as an end

and not simply as a means.

Generally speaking, then, you treat others as ends and not simply as

means when you act on maxims that they, too, if they are reasonable in

their judgments, would judge to bemaxims onwhich it is permissible to act.

Their possessing reason guarantees agreement with the maxim on which

you act if it is a maxim on which it is permissible to act, for no matter who

applies the CI procedure to the maxim the result will be the same. To treat

others as ends and not simply as means requires therefore acting on max-

ims that fulfill the conditions of the formula of universal law. In this way

you honor the humanity, the rational nature, of all with whom you deal.

4. The formula of autonomy and the kingdom of ends

During the course of his account of the formula of humanity, Kant notes

that every person attributes absolute value to his or her own existence as a

rational being. Everyone, that is, realizes that his or her powers of reason

are to be honored as having value that nothing else matches. This recogni-

tion of the absolute value inherent in one�s rational nature is subjective,

Kant says, by which he means that one cannot appeal to it as proof of

rational nature�s absolute value. At the same time, Kant maintains that

one�s recognition of this value does not come through experience. In this

respect it differs from the relative value one finds in objects of natural

desire. One finds relative value in such objects upon experiencing them

and thus seeing in the enjoyment this experience brings how these objects

can contribute to one�s happiness. Or alternatively, on the basis of a desire

for some object, one projects onto the future such an experience and then

infers from the resulting prospective enjoyment of the object how it can


contribute to one�s happiness. In either case the value is discovered through

experience of the world and its impact on one�s well-being. By contrast,

Kant denies that the value of rational nature is discovered in any such

experience of the world. Rather one sees it as the result of reflection on

the limits that categorical imperatives place on the pursuit of the ends on

which one has staked one�s happiness. The limits reveal the authority of

these imperatives. Hence, they reveal the authority of the principle, the

formula of universal law, and the procedure that it defines and that yields

them. The procedure being a form of practical reason, it follows that in the

revelation of its authority the inherent value of one�s rational nature is

revealed as well. From these observations Kant draws the idea of reason�s

being the sole, authoritative source of categorical imperatives, which is to

say, the maker of universal law. This idea is the most fundamental idea of

his theory.

Kant, then, identifies it with a third formulation of the Categorical

Imperative. This formulation, he declares, is a synthesis of the other two.

It resembles thefirstmore than the second, but unlike thefirst, it represents

the universal law that one constructs out of one�s maxim and accepts when

the construction passes the two tests of the CI procedure as a law that one

issues to oneself through an exercise of one�s reason. And given Kant�s

identification of the will with reason,6 the third formulation becomes: Act

only from a will that makes universal law through its maxims.7 The formulation

expresses an ideal of self-government. It is the ideal of one�s actions being

governed exclusively by laws that one has given to oneself. It is, in other

words, the ideal of being autonomous, of being an agent who is subject to

no laws but those he himself has authored. Accordingly, Kant introduces the

phrase �autonomy of the will� to express the idea with which he identifies

this formulation.8 And the formulation itself is commonly known as the

formula of autonomy.

In reaching this formula, Kant believes he has found the explanation of

themotive to act lawfully, themotive of duty, as he sometimes calls it, from

which lawful actions must spring if every such action has a rational basis.

Lawful actions that lack a rational basis are possible when the laws are

externally imposed on their subjects. Examples of externally imposed laws

are laws that kings impose on their subjects and laws that colonial rulers

6 Ibid., p. 412. 7 Ibid., pp. 431–32. 8 Ibid., p. 433.


impose on theirs. In these examples, the subjects require incentives to obey

the law, particularly, in circumstances in which, without the incentives,

they could advance their interests by disobeying but not by obeying. Or in

other words, without these incentives, there is nothing to insure that some-

one subject to externally imposed laws will have in such circumstances a

rational basis for obeying them. Specifically, without the incentives, the

subject�s being rational is no guarantee that his obeying the laws in such

circumstances is not baseless. Matters are different, however, if the laws are

self-imposed as the formula of autonomy requires. If they are self-imposed

in this way, if they are laws that a person gives to himself by making his

maxim a universal law, Kant argues, then there are no circumstances in

which incentives are necessary for providing a rational basis to obey them.

For such laws express a will formed by reason alone, and it is therefore

unnecessary to suppose of a personwhosewill is so formed that he needs an

incentive to act as the law requires. That it is his will is alone sufficient. It

means that when he acts lawfully, he acts without regard to external

influences and consequently, unlike actions that obey laws another has

imposed on him, he acts on initiative that does not originate in the will of

another. He acts, that is, from a motive that originates in his own will.

To help illuminate his idea of the autonomy of the will, Kant introduces

the idea of a kingdom of ends. The former is the idea of the will of a rational

agent as making, through its maxims, universal law. And the idea of a

universal law made through such a maxim is the idea of a law to which

every rational agent, insofar as he or she might disobey it, is subject.9 Since

the idea of the will of a rational agent making universal laws through its

maxims applies to every rational agent, it is possible, then, to think of all

rational agents as being joined together under laws made through the

maxims of any one of them. And since the identity of the agent makes no

difference to what laws would be made through his or her maxims, this

thought is essentially the thought of all rational agents as members of a

legislature that makes laws to which each is subject insofar as he or she

might disobey them. It is the thought, in other words, of a community of all

rational agents governed by laws that they give to themselves collectively.

9 The qualification is necessary because Kant includes God in the category of rational

agents along with human beings. God, unlike human beings, is not subject to law

because he can never will to act contrary to law. Hence, while both God and man are

makers of moral law, only human beings are subject to it.


Such a community is what Kant has in mind in speaking of a kingdom of

ends. It is a kingdom, he says, in being �a systematic union of different

rational beings under common laws.�10 And it is a kingdomof ends, because

the laws join together different rational agents simply as ends in themselves

and therefore without regard to their different identities or to the sort of

personal ends they pursue in their lives. In short, it is a community whose

laws are enacted by the members collectively and are such that the mem-

bers through compliance with them treat each other as ends and never

simply as means. When one acts from a will that regards itself as making

universal law through its maxim, one acts as if one were a lawmaking

member of such a community.

The idea of being a lawmaking member of a kingdom of ends appears to

be Kant�s version of the deontologists� ideal of living in fellowship with

others as one�s equals under the rule of law. For it, too, is an ideal of

belonging to a community in which all men and women are equally subject

to law and equally protected by it. Yet Kant, in constructing this ideal of

egalitarian fellowship, goes further than his deontological predecessors in

the natural-law tradition. For them equality within the fellowship meant

that all were equally subjects of the law that governs and protects them.

Kant, by contrast, made the fellowship explicitly democratic. In his version,

all men and women are equally and collectively the makers of the laws that

govern and protect them. Moreover, his ideal has an importantly different

place in his theory than the corresponding ideal had in those earlier deon-

tological theories.

The difference follows from the difference we noted earlier between

Kant�s conception of law and that of those earlier theories. On the earlier

theories, recall, laws are instruments for maintaining peaceful and stable

social relations. They are needed because human beings cannot live

together peacefully without the imposition of a moral order on their soci-

ety, and such an order is established when they command sufficient obedi-

ence from the society�smembers. More generally, they serve to promote the

common good, and obedience to them, when not due to self-interested

motives, such as fear of the punishment inflicted for disobedience, is due

tomotives of public spirit. Plainly, this conception ofmoral laws goes against

Kant�s conception of them as commanding obedience independently of any

10 Kant, Groundwork of the Metaphysics of Morals, p. 433.


social benefits that obedience might bring. For Kant moral laws command

obedience because reason requires unconditionally that one act as they

command. By so acting one fully realizes one�s nature as a rational being.

One acts as an autonomous agent and not as an agent whose behavior is

subject to the arbitrary influences of natural desire and animal emotion. It is

the full realization of one�s rational nature, then, and not social peace or

other benefits accruing to one�s society, that for Kant explains the place of

morality in human existence. And being a lawmakingmember of a kingdom

of ends is thus, in Kant�s theory, only incidentally an ideal of living harmo-

niously with others as one�s equals. Its principal place in the theory is to

enrich our understanding of what it means fully to realize our rational

nature.

5. Answering the charge of excessive formalism

The deeper theory of morality Kant expounds, beginning with his prelimi-

nary reflections on the value realized through acting lawfully, has the

burden of answering the charge of excessive formalism that is commonly

made against Kant�s ethics. The answer the theory gives should now be

evident. The value one realizes through acting lawfully is that of autonomy

or the full realization of one�s rational nature. Lawful action, in otherwords,

is the same as autonomous action. Consequently, lawful action as such has a

rational basis. For the laws obedience to which qualifies one�s actions as

lawful are laws that one gives to oneself, and such laws are in every case the

enactments of practical reason. The charge of excessive formalism, then,

according to this answer, represents an incomplete understanding of lawful

action. When the charge is made, it is made against Kant�s ethics as defined

by the formula of universal law. Kant�s deeper theory of morality is not

considered. Hence, that the agent is the maker of the laws obedience to

which qualifies his actions as lawful goes unrecognized. But once lawful

action is understood to be the same as autonomous action, action through

which one fully realizes one�s rational nature, the seeming pointlessness of

acting lawfully in the examples on which the charge is based disappears. So

too does the charge. Or at least that is the answer Kant�s theory gives.

What this answer implies is that we should not expect to find in Kant�s

deeper theory of morality a way to correct the application of the CI proce-

dure in the cases that yield false negatives so that the procedure yieldsmore


acceptable conclusions. Rather, we should expect to get from the theory a

defense of the conclusions the procedure delivers in those cases. That is, we

should expect to get from it a defense of the conclusions that up to now we

have described as false negatives. These are, in each case, a conclusion that

to act on the maxim in question would be to act unlawfully. Specifically,

each is the result of one�s having determined, by applying the CI procedure,

that the maxim is unfit to be made, through one�s will, a universal law.

Formally, its unfitness is seen in the contradiction one finds either in one�s

conception of the maxim as a universal law or in one�s will when one

regards it as becoming a universal law through one�s will. But these formal

criteria shed no light on why one should regard acting on it as wrong. They

indicate only that reason is confounded when one considers whether the

maxim can become through one�s will a universal law. Hence, Kant�s ethics

appears excessively formalistic when one looks only to these criteria for

explanation of why acting on thismaxim is wrong. Onemust therefore look

to the other formulations of the Categorical Imperative and the criteria by

which they determine that themaxim is unfit for beingmade through one�s

will a universal law to see why acting on it is wrong. And in this regard,

looking to the third formulation, the formula of autonomy, is particularly

illuminating when its core idea of a will that regards itself as making law

through its maxim is interpreted as that of being a lawmaking member of a

kingdom of ends.

When the formula is so interpreted, its test for whether your maxim is

fit to be made a universal law becomes whether you could, as a lawmaking

member of a kingdom of ends, enact it as a law. All such enactments

necessarily proceed without regard to anyone�s identity or special inter-

ests, for lawmaking in a kingdom of ends must be impartial. If it were not,

then there would be laws designed to work to the benefit of some and at

the expense of others, and the existence of such laws is contrary to the

ideal of a kingdom in which all, being ends in themselves, are always

treated as such. After all, a law designed to benefit some at the expense

of others could not be the product of the will of any of the latter, since

whatever maxim was the basis of its enactment could not be one that any

of them found reasonable. It would not satisfy the CI procedure. It follows

that a maxim action on which would achieve its end only if the agent is

alone or among a small number in so acting is unfit for enactment as a law

in a kingdom ends. The reason is that the benefits of acting on this maxim


cannot be shared by all who act on it should everyone choose to do so.

Hence, because the benefits are available to those who act on it only if a

limited number do so, it makes sense to act on the maxim only if one can

be confident that enough people will refrain from so acting as to enable

one to gain the benefit. In other words, it makes sense to act on themaxim

only when one can take advantage of their refraining from acting on it. In

taking this advantage, however, one would be treating oneself as different

from others, and to do so would be inconsistent with the impartiality

toward all that acting only onmaxims that can be made laws in a kingdom

of ends entails. The wrongfulness of the action, then, consists in the

favoritism toward oneself implied in one�s acting on a maxim that one

could not will that everyone act on.

To illustrate the point, consider again the example of your carrying a

handgun for protection in circumstances in which an increase in the inci-

dence of violent crime in your neighborhood has made you concerned

about going out at night. This action, as we noted before, makes sense as

long as only relatively few people carry handguns, for if everyone carried

one, the risk of your being harmed by violence would increase rather than

decrease, notwithstanding the protection your carrying one provides. You

would therefore be acting on a maxim whose end is achievable only if the

number of people who act on it is limited, which is to say that the greater

personal safety you would gain by acting on this maxim is not available to

all who act on it should everyone do so.11 By acting on thismaxim, then, you

would be advancing your interest in personal safety in a way that you could

not will that everyone take advantage of. Consequently, you would be

displaying an attitude of favoritism toward yourself since the action

would express your willingness to seek a benefit for yourself that you

would not will that everyone seek for him- or herself in the same way.

Such an attitude is inconsistent with the impartiality toward all required

of every lawmaking member of a kingdom of ends. It expresses, in Kant�s

view, a lack of respect for others as ends in themselves. Thus one can

explain why the action is wrong in a way that is intuitively more satisfying

than citing the failure of its maxim to satisfy the formal criteria necessary

for it to become a universal law through the agent�s will.

11 For the statement of the maxim (N2) on which you would be acting, see above, p. 154.


6. Rationalism revisited

Of course, this explanation, though intuitively more satisfying, may still

leave doubters. It will not convince everyone that it would be wrong for

someone concerned about the rise of violent crime in his neighborhood to

carry a handgun for protection. If you are among the unconvinced, then you

might still think that Kant�s ethics is unsound. And one reason why is that

you think his account of morality is too austere. Morality, you might think,

does not prohibit people from carrying handguns for protection in circum-

stances in which there is a heightened risk of their being victims of violent

crime. And more generally, it does not require us to forgo advantages that

come from actions that we know many people will not do and that are

advantageous so long as this is true. Kant then, on this view, simply errs in

holding the contrary. Nonetheless, you might still wonder whether this

view is really sufficient to prove Kant�s ethics unsound. If Kant�s ideal of a

kingdom of ends in which all rational beings are legislative members

uncovers something aboutmorality of which youwere previously unaware,

perhaps there is room to weigh the certitude with which you hold these

moral convictions that contradict the conclusions of the CI procedure

against the attractiveness of this ideal and its support for those conclusions.

This possibility should seem especially strong if you have no competing

ideal to which you could appeal to support and unify your convictions.

Be this as it may, there is a second reason why you might still think Kant�s

ethics is unsound. For not only are some of the conclusions the CI procedure

yields contrary to common-sense judgments about what we are free to do and

whatwe have a duty to do, they also imply that to act contrary to them is to act

against reason, and the idea that a person acts unreasonably by, for example,

carrying a handgun to protect himself from violent crime in circumstances in

which such crime in his neighborhood is on the rise is, if anything, much

harder to fathom than the idea that he violates a duty by so acting. What,

after all, couldbemore rational thanone�s taking suitablemeans to achieving a

reasonable personal end, an end on the achievement of which one has staked

one�s happiness? Accordingly, you may find, in questioning the credibility of

Kant�s bold thesis that obedience tomoral law is essential to rational action, an

evenmore powerful reason to think his ethics unsound. Indeed, this challenge

to Kant�s rationalismwill prove to be the greater threat. Ultimately, as we will

see, it reveals a serious gap in his system.


At the core of Kant�s rationalism is the thesis that moral laws are catego-

rical imperatives. What Kant understands by this is that obedience to moral

law is obedience to an unconditional dictate of reason. The chief corollary of

this is that one cannot disobey a moral law without going against reason.

And this corollary is especially striking given Kant�s austere conception of

morality. For on that conception, as we�ve seen, a person who seeks advan-

tages for himself by acting on amaxim that he cannot will that everyone act

on is not only in breach of a moral duty but also acting unreasonably, and

this latter proposition defies belief. To be sure, the person acts on a maxim

that he can neither conceive of as a universal law nor will that it become

one, but this observation only moves our skepticism about Kant�s ration-

alism back one step. Why, after all, is it unreasonable to act on such a

maxim? The problem is this. Whether or not the CI procedure correctly

tracks the reasoning that yields conclusions about what our duties are, the

question remains why rational deliberation requires that these conclusions

supersede conclusions about how best to achieve our personal ends. They

would, of course, if they were categorical imperatives, for reason requires

that one follow a categorical imperative whenever its directive conflicts

with the directive of some hypothetical imperative.12 But one cannot credit

the conclusions of the CI procedure with being categorical imperatives just

in virtue of their being the conclusions of this procedure. For a categorical

imperative is a dictate of reason, and as we�ve seen, neither the procedure

nor the formula of universal law that the procedure implements alone

provides a rational basis for following its conclusions. So to resolve the

problem we need to look to Kant�s deeper theory of morality. It is there

12 Reason requires this because categorical imperatives are unconditional dictates of

reason and hypothetical imperatives are conditional dictates of reason, the condition

on which each dictates being the agent�s having willed a certain end. Thus if the agent,

in deliberating, sees that he cannot follow a hypothetical imperativewithout ignoring a

categorical one, reason requires that he follow the latter. He must follow the latter, for

in doing so he avoids violating either dictate of reason, which is to say he avoids acting

against reason. Specifically, he does not violate the hypothetical imperative when he

follows the categorical one but rather invalidates it by ceasing to will its validating end

(i.e., the end the necessarymeans towhich the imperative directs him to take). In effect,

then, when he follows the categorical imperative, he removes the condition on which

the hypothetical imperative is valid and so avoids violating it. Were he to follow the

hypothetical imperative instead, he would violate the categorical imperative, which is

to say, he would go against reason. Hence, reason prohibits him from following a

hypothetical imperative when doing so means ignoring a categorical one.


that we find Kant�s answer to the question of why rational deliberation

requires that the conclusions of the CI procedure supersede conclusions

about how best to achieve one�s personal ends.

The answer would appear to fall directly out of a comparison of the

rational bases for following the two types of imperative. Recall that the

rational basis for following a categorical imperative is the value one realizes

through acting lawfully. This is the value of acting autonomously, and being

inherent in rational nature and therefore a value for every rational being, it

is absolute and constant across different rational beings. In both respects it

differs from the value that explains the rational basis for following a hypo-

thetical imperative. This is the value of achieving the personal end the

necessary means to which the imperative directs one to take. Achieving

this end has value by virtue of its contribution to one�s happiness, and

therefore different actions and conditions have value for different people

depending on the natural desires and personal interests whose satisfaction

constitutes their happiness. Accordingly, the value of achieving a personal

end is relative, and moreover its magnitude varies according to how great a

contribution the achievement makes to one�s happiness. Yet however great

the achievement�s contribution, its value, according to Kant, cannot be

compared to the value of acting autonomously. To the contrary, the latter,

being absolute, limits what one can do in pursuit of ends whose achieve-

ment has the former. There is, then, no basis for trade-offs between

autonomy and the means for achieving one�s personal ends. Rather one

must choose the latter within the constraints set by the respect due the

former. Conformity to categorical imperatives, in other words, constrains

what hypothetical imperatives one can follow. And this constraint means

that one cannot ignore a categorical imperative in order to follow a hypo-

thetical one.

Plainly, Kant in this answer rests his conclusion on his distinction

between the absolute value of autonomous action and the relative value

of achieving a personal end. Yet the distinction cannot by itself support the

conclusion. For the distinction does not imply that one type of value is

superior to the other. It does not imply that an absolute value is superior

to a relative one. Rather, to say that something has absolute value is to say

that its having that value does not depend on the existence of anything

contingently related to it, whereas to say that something has relative value

is to say that its having that value depends on the existence of something


that is contingently related to it. In the case at hand, autonomy has value

and there are no conditions under which it could fail to have it. By contrast,

achieving a personal end has value only so long as it satisfies some desire or

interest the object of which is that end. For only then does it contribute to

one�s happiness. Hence, since the end�s being the object of some desire or

interest is a contingent relation, since it is possible, that is, for one to will an

end without that end�s being the object of any desire or interest, there are

conditions under which the action could fail to have value. Specifically, if

the desire or interest whose object is the end one has achieved did not exist

and the end was not the object of some other desire or interest, then

achieving the end would not have value. This possibility, however, though

it explainswhy the value of autonomy is absolute and the value of achieving

a personal end is relative, does not support regarding either value as supe-

rior to the other.13 After all, while the achievement�s having value depends

on the existence of a desire or interest whose object is the end one has

achieved, there is nothing in this dependence to imply that the value it has

is inferior to the value of autonomy.

Kant believes otherwise. The soaring language by which he characterizes

the difference between the value of autonomy and that of achieving a

personal end implies as much. Thus, having identified the idea of relative

value with that of a price, he writes, �[T]hat which constitutes the sole

condition under which anything can be an end in itself has not merely a

relative value – that is, a price – but has an intrinsic value – that is, dignity.�14

And he further characterizes the dignity that ends in themselves have as

�exalted above all price.�15 Similarly, he attributes to the mental attitude

distinctive of autonomous action �an intrinsic worth . . . [that] puts it

infinitely above all price, with which it cannot be brought into reckoning

or comparison without, as it were, a profanation of its sanctity.�16 Finally,

distinguishing human beings as free when the laws that govern their action

are laws that they give to themselves and unfree when the laws that govern

their action are the laws of nature that regulate the stimulation of their

13 At one point Kant writes, �[I]f all value were conditioned – that is, contingent – then no

supreme principle could be found for reason at all� (Groundwork of the Metaphysics of

Morals, p. 429), from which one may be tempted to infer that the existence of absolute

value implies that there is a supreme principle. But the inference in that case would be

faulty, an instance of the fallacy known as denying the antecedent.14 Ibid., p. 435. 15 Ibid., p. 434. 16 Ibid., p. 435.


appetites and passions, he characterizes the lawmaking essential to this

ideal of human freedom as having �a dignity – that is, an unconditioned and

incomparable worth – for the appreciation of which, as necessarily given by

a rational being, the word �reverence� is the only becoming expression.

Autonomy,� Kant concludes, �is therefore the ground of the dignity of

human nature and of every rational being.�17

But the distinction between dignity and price to which Kant appeals in

these passages to capture the distinction between the value of autonomy

and the value of achieving a personal end is a different one from the

distinction between an absolute value and a relative one to which he first

appealed in distinguishing the value of autonomy from the value of achiev-

ing a personal end. For the distinction between dignity and price includes a

ranking of one value over the other, and as we already noted, nothing in the

distinction between absolute and relative value supports taking one type as

superior to the other. Consequently, the distinction between absolute and

relative value does not support the thesis Kant intends when he attributes

dignity to autonomous action and price to achieving a personal end. This is

the thesis that the value of the former is incomparably superior to the value

of the latter. And unfortunately for Kant, without it he cannotmaintain that

the conclusions of the CI procedure are categorical imperatives and so

cannotmaintain that ignoring any of them in order to follow a hypothetical

imperative is never reasonable. His shift to the language of dignity, rever-

ence, and infinite value helps, then, to obscure the gap in his argument for

this conclusion. It does not close it.

What leads Kant to shift to this language? Why does he think he has

shown that the value of autonomy is incomparably superior to the value of

achieving personal ends? The answer must lie in his view of human beings

as having a dual nature. At the same time, it should be evident that the

duality of human nature alone is insufficient to explain why the value of

autonomy is the superior value. Atmost it could explain why the two values

are incommensurable. Their being incommensurable, however, merely

means that comparing the worth of acting autonomously with the worth

of achieving a personal end is like comparing the worth of a well-performed

ballet with that of well-played violin concerto. While individuals will differ

about which performance they prefer – ballet enthusiasts will prefer the

17 Ibid., p. 436.


former, devotees of classical music the latter, there is no grounds for taking

one performance to be superior to the other absolutely. So if the answer lies

in Kant�s view of human beings as having a dual nature, that view must

involve more than the mere duality of our nature.

What more it involves is the chief doctrines of rationalism: that the

sensuous aspect of human nature gives rise to motives of desire and emo-

tion that conflict with motives originating in reason and that the values

implied by the latter must be supreme in a human life else that life will be

given to unreasonable pursuits of pleasure and unreasonable abandon-

ments of morally worthy ends. For Kant, therefore, any conflict between a

conclusion of the CI procedure and a hypothetical imperative that directs

one to take necessary means to the achievement of a personal end can be

understood as a conflict between reason and some desire or emotion that

arises from the sensuous aspect of one�s nature. At the same time, this

understanding of such conflicts requires recourse to his deeper theory of

morality to ground the CI procedure�s conclusions in reason. But this

grounding, as we have seen, is insufficient to support regarding them as

superseding in deliberation any conclusion about how best to achieve a

personal end and so superseding any hypothetical imperative that directs

one to take the necessarymeans to such an end. So ultimately Kantmust fall

back on the very doctrines of rationalism his theory is meant to uphold to

maintain that the conclusions of the CI procedure supersede in deliberation

conclusions about how best to achieve one�s personal ends. The doubts

about the soundness of Kant�s ethics that arise from resistance to the thesis

that there is some defect in your reason if you choose to carry a handgun for

protection despite its being contrary to the CI procedure reflect doubts

about these rationalist doctrines. Kant�s ethics, one could say, is a pro-

foundlymodern elaboration of the doctrines. But it falls short of vindicating

them.

7. Personal autonomy

Part of the profundity of Kant�s ethics is its use of the theme of self-

government to explain morality�s authority. To propose that men and

women are the authors of the moral laws that govern their lives is to

break completely from traditional conceptions of morality as externally

imposed on human beings or as implanted in their frame by nature or


some other external force. Seventeenth- and eighteenth-century political

thought had produced various ideals of republican freedom according to

which, on the one hand, no adult human being of sound mind and intact

reasonwas subject by nature to the rule of any otherman or woman and, on

the other, political rulers, towhose authoritymen andwomenwere subject,

ruled by their consent. Themost progressive strains in this thought not only

founded the authority of governments on such consent but also made

governments subject to democratic control. An ideal of popular sover-

eignty, then, came to represent the most forward-looking programs in the

political philosophy of the late eighteenth century. And Kant�s genius as a

moral philosopher lay in his seizing on this ideal and extending it to

morality. The kingdom of ends is nothing less than an ideal of popular

sovereignty in which the people are all rational beings and the society

over which they are sovereign is a fictional union of them. Thus Kant

extended the spirit of emancipation from feudal and hierarchical social

institutions that animated modern political thought to the fixed moral

order that, on traditional conceptions of morality, structured the lives of

human beings. Morality, on Kant�s view, ceased to comprise standards

whose authority one was subject to as a servant is subject to the authority

of the master�s rules or a trainee is subject to the authority of a trainer�s

regimen. Rather being the author of these standards, one obeyed them as

one obeys one�s own freely made decisions about how to act. Obedience to

them, on this view, is a form of self-determination.

This idea that each person is the author of the moral standards that

govern his or her life has had a powerful hold on philosophical thought

even among Kant�s sharpest critics. For one can reject his rationalism and

the austere conception ofmorality that goes with it and still retain this idea.

It is consistent, for instance, with the view that, facedwith a choice between

following a conclusion of the CI procedure that directs one to forgo advan-

tages one could gain as long as few others did so as well and pursuing a

personal end by taking those advantages, one does nothing wrong in choos-

ing the latter. To choose the lattermay require that one abandon the ideal of

being a lawmaking member of a kingdom of ends, but one need not uphold

this ideal to conceive of oneself as the author of the standards one follows. It

is sufficient if one has freely, reflectively, and sincerely decided to live

according to certain standards and not others. The point is that Kant�s

notion of the autonomy of the will is not the only notion of individual


self-determination available to moral philosophers once they entertain the

idea that the moral standards a person lives by are themselves determined

by that person�s own choices. And a number of such philosophers in the two

centuries since Kant published his great works have realized this point

and advanced alternative notions and a corresponding ethics of self-

determination based on them. The term �existentialism� is often used to

refer to the movement of thought they represent, though Jean-Paul Sartre

(1905–80) may well be the only major philosopher among them to have

embraced the term. Nonetheless, I will follow this usage.

The starting point of existentialism is human thought. Unlike Kant,

however, existentialists do not focus on the powers of reason in character-

izing what is distinctive about human thought. Rather they focus on the

powers of decision and choice. What makes these powers distinctive of

human thought is that their exercise typically resolves matters presented

to the mind. Someone, for instance, finds himself in a situation in which he

hears screaming of an uncertain kind that is coming from a house nearby,

and hemust decide whether the screams he is hearing are those of a child in

need of help or those of a child engaged in exuberant play. Similarly,

someone finds himself in a situation in which, at a time when he is rather

busy, he is asked by a stranger for assistance and offered payment in return,

and he must decide whether the payment would be worth the time and

effort that assisting the stranger requires. In these examples, the decision is

about some matter in the world external to the thinker�s mind that is

presented to him, screaming in the first example, an offer from a stranger

in the second, and he must decide how to resolve it because what is pre-

sented to him is ambiguous or uncertain. Resolution, in other words,

depends on his judging on the basis of the evidence or considerations before

him rather than on his merely responding to the force of the different

images and sensations that the world impresses on his mind. While the

latter may characterize how such ambiguity or uncertainty is resolved in

the thought of nonhuman animals, the former is, arguably, distinctive of

human thought. The famous plight of Buridan�s ass, immobilized by being

situated at equal distance from two equally large and equally appetizing

piles of hay, illustrates the point.

This contrast highlights the difference between an active mind and a

passive one. It is a major theme of existentialist thought that human beliefs

are, far more often than is typically acknowledged, the product of the


mind�s being active in their formation, which is to say that human beliefs

result, far more often than is typically acknowledged, from decisions like

the ones in these examples. Recognizing that a belief one holds resulted

from such a decision means also recognizing a certain degree of personal

responsibility that one has for holding the belief. A second major theme of

existentialist thought, then, is that men and women are far more respon-

sible for many of the beliefs they hold and the states of mind that flow from

them than is typically acknowledged. And because accepting responsibility

implies assuming a kind of burden, which many people would prefer to

avoid, a thirdmajor theme of existentialism is the flight from accepting this

responsibility that is common to human life. The reason therefore why

people seldom acknowledge the role of their own decisions in the forma-

tion of their beliefs is that they wish to be free of responsibility for those

beliefs. This wish, in turn, gives rise to a strong tendency to regard their

beliefs as formed in direct response to the sensations and images that

impress their minds and other external forces with respect to which they

are also passive. And some of the most celebrated existentialist writing

consists of portrayals of the different ways people deny and hide from

themselves their own complicity in the forming of beliefs central to their

outlook on life.

In addition to decisions about matters in the world external to the

thinker�s mind, a person also makes decisions about matters internal to

his mind. Someone, for instance, finds himself in a situation in which he

feels threatened or anxious, and he must decide whether the object that is

the source of his apprehensiveness is properly to be feared and shrunk from

or faced with confidence and challenged. Similarly, someone finds himself

in a situation in which he is invited to do something illicit, and troubled by

his conscience, he must decide whether to listen to it and heed its warnings

and reproaches or ignore it and act on whatever desires the illicit action

appeals to. Or again someone faces a situation in which remaining loyal to a

friend means jeopardizing a mission of national importance, and he must

decide what matters to him more, friendship or the public good. In these

examples, the decision is about somematter interior to the thinker�s mind,

what emotion to feel, what motive to act on, what most to care about, and

hemust decide how to resolve it because in each case it is a question of what

sort of person to be, a question that, according to existentialism, is always

and necessarily open. It is always and necessarily open, for human beings


would not otherwise be capable of the kind individual self-determination

that existentialism attributes to them. In short, theywould not be capable of

what I�ll call personal autonomy.

The decisions in these examples represent exercises of personal

autonomy. As such, they presuppose capacities for self-reflection and self-

assessment. They are, therefore, if anything, even more indicative of dis-

tinctively human thought than the decisions in the earlier examples. The

minds of nonhuman animals respond to the images and sensations pre-

sented to them, and these responses – the whetting of appetite, the stirring

of emotion – define the significance of those images and sensations in

animal life. What is distinctive of human minds are the capacities to take

such responses and other mental states as objects of reflection and evalua-

tion. For self-assessment consists in large part in the evaluation of the ends

that define one�s life pursuits and themotives predominant in howone lives

one�s life, and the evaluation of ends andmotives is the evaluation of states

of mind that contain one�s purposes and move one to act. While a good

watchdog responds with anger and prepares to attack when it sees a

stranger trespass on the property it guards, the dog never reflects on this

anger or wonders whether it is proper to feel it at the sight of a stranger. No

watchdog has ever been known to decide to bemore gentle in its disposition

towards strangers, to greet themwith a wag of its tail rather than a growl or

a bark. Nor has any been known to have brought about such changes in its

demeanor on its own. Human beings, by contrast, are capable of question-

ing their emotions and motives, assessing their importance, and deciding

whether or not to maintain and act on them. This is the essence of their

personal autonomy. And however reluctant people may be to exercise it,

this capacity implies a far greater degree of responsibility for who they are

than is typically acknowledged. This attribution of responsibility to human

beings is at the core of existentialist ethics. And the three major themes of

existentialist thought mentioned above extend to these decisions as well.

8. Existentialist ethics

When someone hears a child screaming and decides that the screaming is a

cry for help, his decision can, at least in principle, be checked against the

facts. And if, as amatter of fact, the screamshe hears express the excitement

of children at play and are not cries for help, then his decision is mistaken.


Similarly, when someone is asked for assistance and promised payment in

return, his decision as to whether such assistance is worth the time and

effort it entails can, at least in principle, be judged with respect to howwell

it fits in with the more general plans and values around which he has

organized his life. And if, as a matter of fact, his assisting the stranger and

receiving payment in return does more to thwart those plans and to betray

those values than to forward and promote them, then his decision is a

mistake.

By contrast, when someone finds himself in a situation in which remain-

ing loyal to a friend will jeopardize amission of national importance and he

must decide whether friendship or the public good should ultimately mat-

ter more to him, there are no facts against which the decision can be

checked. Some people in this situation would, no doubt, put patriotism

ahead of friendship and choose to betray a friend in order to protect the

mission. But there are, with equal certainty, others whose choice would

correspond to E.M. Forster�s memorable statement, �If I had to choose

between betraying my country and betraying my friend, I hope I should

have the guts to betraymy country.�18 And one cannot say of those in either

group that their decision is the correct one and that the decision of those in

the other group is not. If the facts are the same in either case, then they do

not resolve the question of which should matter more, loyalty to a friend or

loyalty to country. It is up to the individual to decide.

Nor is there a standard that all who deliberate soundly must apply to the

situation and that resolves this question. One�s deliberation can, of course,

be more or less sound, but the standards one applies in deliberating are not

among the factors that make it sound. Rather what make it sound are the

clarity and cogency of the thinking, the correctness of the understanding of

the situation, and the sincerity with which some considerations are given

more weight than others. Decisions like this one, then, decisions about

what sort of person to be and by what values and principles one should

lead one�s life, while they may be more or less soundly made, are not of a

kind that can still bewrong if soundlymade. They are decisions inwhich the

discretion to choose one way or the other is unbounded. They represent

radical choices, choices for which there are no criteria of correctness

beyond their resulting from sound deliberation. And the existentialist

18 E.M. Forster, Two Cheers for Democracy (New York: Harcourt Brace, 1951), p. 68.


doctrine of radical choice is that occasions for making such choices are

essential to human life, however much people may try to avoid acknowl-

edging them. In making these choices one exercises one�s capacity for

personal autonomy, and in avoiding acknowledgment of an occasion for

making a radical choice, one hides from oneself the fact of one�s having this

capacity. This doctrine challenges directly the very idea of an ethical theory.

Indeed, it suggests that the enterprise of ethical theorizing is yet another

stratagem for avoiding acknowledging one�s capacity for autonomous

choice.

Sartre, in a famous public lecture on existentialism, pressed this chal-

lenge to ethical theory with a detailed example of a student of his who,

during World War II, while the Germans occupied France, came to him for

advice.19 The student yearned to join the Free French Forces in England. His

older brother had been killed fighting the enemy during the 1940 invasion,

and wanting to avenge his brother�s death, the student felt the need to join

in the struggle to rid his country of her Nazi overlords. At the same time, he

was living alone with his mother, who had separated from his father

because of the latter�s inclination to collaborate with the enemy, and she

was wholly dependent on him. His departure, he knew, would cause her

enormous hardship, aggravating her grief over the loss of her eldest son and

leading to crippling despair. What is more, realizing his aim of joining the

Free French and advancing the cause of liberating France was only a remote

possibility, given all that could interfere with his plans, whereas he was

certain to be of help to his mother on a daily basis if he remained at home

with her. This student�s quandary, Sartre argued, demonstrates the futility

of relying on ethical theories for help in deciding what one should do. The

standard of neighbor love in Christian ethics, for example, is unhelpful; for

it does not say whether neighbor love is shownmore in acts of patriotism or

in acts of filial duty. And Kant�s formula of humanity is equally unhelpful

and for similar reasons. Indeed, any theory, Sartre maintained, is bound to

fall short of providing a satisfactory answer. It ismerely a set of abstractions,

and concrete situations like his student�s necessarily outrun them.

Ultimately, then, it is up to the student to decide which course of action is

19 Jean-Paul Sartre, �L�existentialisme est un humanisme�; reprinted as �Existentialism Is a

Humanism,� Philip Mairet, trans., in Existentialism from Dostoevsky to Sartre, Walter

Kaufmann, ed. (Cleveland: TheWorld Publishing Co., 1956), pp. 287–311, esp. pp. 295–96.


the right one, and there is no way for him to determine this independently

of his making that decision. As Sartre put it, the student does not find the

right answer to the question of what he should do; he invents it.

This last point highlights the existentialist view of human beings as

creators of the values by which they lead their lives. It captures their under-

standing of the capacity for personal autonomy, just as Kant�s idea of one�s

being the maker of the moral laws that govern one�s life captures his

understanding of the autonomy of the will. And just as the latter yielded

the idea of one�s being a lawmaking member of a democratic republic

whose citizenry was all rational beings, so the former yields the idea of a

solitary artist who creates value through the artistic endeavors in which he

or she engages. Sartre, in particular, finds the similarity between the artist�s

decisions and those ofmoral life ethically instructive. The point, however, is

not to compare moral value to aesthetic value. Rather it is to compare the

creative acts of artists, for which there is no question of their being respon-

sible, to actions in human life that imply decisions concerning the sort of

person to be or the things that should ultimately matter to one. By implica-

tion, then, a person who relies on doctrines and theories in ethics in decid-

ing how to live is like someone who paints by the numbers or follows the

instructions in a manual for would-be painters. In either case, the person

avoids exercising his or her own distinctively creative powers and substi-

tutes instead something mechanical or automatic. He does not want to

acknowledge that what is done or what is put on the canvas is something

of which he is the author and for which he is thereby responsible. He wants

instead to shift responsibility to some external authority. He wants, that is,

not to have the freedomhe in fact has to decide how to live his life orwhat to

display on the canvas before him.

Existentialist ethics takes this freedom, the capacity for personal

autonomy, as the sole condition with respect to which actions are morally

evaluated. The standard it applies, however, is not a universal standard of

right action. Indeed, on existentialist ethics there are no such standards.

The standard it applies, rather, concerns the self-understanding with which

a person exercises this freedom in acting. It is the self-understanding of

someone who wholly identifies with his actions. Such a person recognizes

and accepts his actions as the products of his own decisions, which he

understands in turn as resulting from his exercising his freedom. He under-

stands, then, that he is the author of these actions and as such is responsible


for them. In a word, it is a standard of integrity. One who meets it does not

try to put distance between himself and his actions by regarding them, say,

as caused by alien forces or compelled by external authorities. And he does

not try to avoid responsibility for them by regarding his doing them as

necessitated by his nature, his upbringing, or his place in society.

Someonewho does regard his actions in any of these ways, notwithstanding

their having resulted fromhis deciding to do them or fromhis being passive

in situations in which he could have so decided and acted accordingly,

denies that he has the capacity for personal autonomy. He acts under a

false view of himself as unfree, as having no choice in the matter. The self-

understanding with which he acts is that of someone who deceives himself

about his actions and his responsibility for them and therefore falls short of

the standard of integrity by which the defenders of existentialist ethics

evaluate the morality of actions. In Sartre�s well-known expression, he

acts in bad faith. In existentialist ethics, then, someone acts well when his

or her actions display integrity. He or she acts badly when they are done in

bad faith. And any other judgments about themorality of one�s actions are a

matter of personal decision.

Of what use is this ethics in answering questions about how one ought to

live and what actions ought one to do in the conduct of one�s life? How, for

instance, does it help you to determine what to do in the situation in which

you have found a lost purse containing a huge wad of cash? Plainly, you will

notfind in existentialist ethics an answer to the question ofwhat to do in this

situation. None of its doctrines generates for you a reason for returning the

purse and the cash to its owner, nor does any of them generate for you a

reason for doing something else. Instead, they bid you to see this situation as

one of radical choice. It is a situation in which you must decide what sort of

person to be and in which there are no facts determining what the correct

decision is and no standard that you are required to apply in deliberating

aboutwhat you should do. The only standard to be applied is that of integrity,

and that standard does not enter your deliberations as a standard of right

action. Rather it enters by calling on you to be conscious of the freedom you

have to decide whether to be the sort of person who respects property and

seeks to restore lost valuables you happen to find to their rightful owner or

who disregards property and takes advantage of other people�s losses.

Hence, a decision to keep the cash is a decision to be someone of the

latter sort. If in making it, however, you try to justify keeping the cash on


grounds that implicitly deny the freedom you have to be a different sort of

person, then you fail to meet the existentialist standard of integrity. If you

think, for instance, that your needs of the moment justify your keeping the

cash or that youwere destined tofind it and to use it as you seefit or that it is

a windfall and thus fairly kept by you, etc., you act in bad faith. If, by

contrast, you recognize the decision to keep the cash as an exercise of

your freedom, an exercise in which you freely choose to be someone who

does not respect property and takes advantage of other people�s losses and

do not try to explain your choice as permitted or justified by extenuating

circumstances or some special dispensation, then nothing in existentialist

ethics speaks against your keeping it. It would be an unusual thief, to be

sure, who had such integrity, but if what defines human beings as moral

agents is a capacity for personal autonomy, then the possibility cannot be

gainsaid.

9. The excesses of existentialism

Does attributing the capacity for personal autonomy to human beings have

the striking implications for ethics that existentialism draws from it? Does

the attribution imply, for instance, that there are no universal standards of

right action? Does it imply that there are no values inherent in human life

by which to evaluate the ends people pursue? These implications, if sound,

would represent a powerful objection to ethical theory. But are they sound?

A review of Sartre�s arguments shows that he at least, in drawing these

implications, overstates the existentialist case. Consider,first, his argument

in criticism of ethical theory that he bases on the example of the student

who was torn between joining the Free French Forces and looking after his

mother. Sartre maintains that no ethical theory can resolve the student�s

quandary because the complexity of a concrete situation, such as the

student�s, exceeds what the abstract principles a theory consists of can

comprehend. But the premiss of this argument is mistaken. Utilitarianism

certainly has no difficulty comprehending the situation. Nor does Kant�s

ethics, despite Sartre�s remarks to the contrary.

Utilitarianism has no difficulty because by design the principle it identi-

fies as at the foundation of ethics, the Principle of Utility, comprehends all

practical situations. This point should be obvious, since in any situation in

which a personmust decidewhat to do, nomatter how complex it is, at least


one of the actions open to him will satisfy the utilitarian standard of right

action. At least one of them, that is, will bring about as much or more

happiness overall than any of the other actions open to him. To be sure,

complex situations like the student�s present practical difficulties in carry-

ing out the calculation necessary for determining which action satisfies the

utilitarian standard. But these difficulties are no different in kind from the

difficulties of making such a calculation in less complex circumstances.

After all, one can never be certain of all the consequences of an action one

contemplates doing, and these difficulties simply contribute to the uncer-

tainty that is present in a situation. Sartre, then, though he has described a

situation in which there is a great deal of uncertainty about the outcome of

one of the courses of action a person is contemplating, has not described a

situation to which the utilitarian standard is inapplicable. His student could

still have applied it in trying to resolve his quandary, and while his assess-

ment of the consequences of his following each course of action and their

likelihood would be crude, his making a decision on the basis of that assess-

ment, in accordance with the Principle of Utility, would give it theoretical

support.

So too Kant�s ethics has no difficulty comprehending the student�s sit-

uation. This is because the complexity of the situation does not prevent the

student from identifying either the maxim on which he would act if he

opted for joining the Free French Forces or the maxim on which he would

act if he opted for staying with his mother. Therefore it does not prevent

him from determiningwhether eithermaxim satisfies the two tests that the

CI procedure comprises. To be sure, applying the CI procedure to these

maxims might still leave him undecided about which course of action to

take, since both maxims may satisfy the two tests. But this is not a defect in

Kant�s ethics, for his theory is not designed to yield an answer to every

question about what one ought to do. Rather the principle it identifies as at

the foundation of ethics, the Categorical Imperative, determines whether

acting on a certainmaxim in a given situation is lawful, and in any situation

there may be more than one maxim on which one could act lawfully, in

which case the theory is silent on which maxim to act on. Thus Sartre is

simply mistaken in thinking that Kant�s ethics must be useless if it cannot

resolve his student�s quandary.

Of course, Sartre may think the student�s situation is a crucial test of the

adequacy of an ethical theory because the situation entails a conflict of


obligations and he thinks the adequacy of an ethical theory depends on its

ability to resolve such conflicts. But this thought, too, would be mistaken.

We should not think that every conflict of obligation puts an ethical theory

to the test. It does so only if the person faced with the conflict could act

wrongly by fulfilling one of the conflicting obligations rather than the

other, and not every conflict is like this. Sometimes the conflict is between

two obligations that are specific forms of a more general obligation, and as

long as you fulfill the more general obligation by fulfilling one of these

specific forms, you do nothing wrong by disregarding the other. Those of us

who are prosperous and healthy, for instance, have obligations to help the

needy and to help the sick, yet we shouldn�t expect a theory like Kant�s to

resolve any conflict we have between donating to a food bank and donating

to a hospital. Donating to either would presumably be found lawful if the CI

procedure were applied to the maxim on which we would act, for in so

acting we would be fulfilling our obligation to contribute to the good of the

less fortunate. Consequently, there would be nothing further for a theory

like Kant�s to say. And if the situation Sartre�s student faced is similarly one

of conflicting obligations either of which the student could fulfill without

acting wrongly, then the situation does not provide a test of the theory�s

adequacy.

Sartre�s argument fails, then, to show that ethical theories are incapable

of providing guidance in complex situations. There is, however, a different

criticism of them he could have made. He could have criticized them as

falsely offering us safe havens from having to make radical choices. This is

the signature criticism of ethical theories that is implicit in existentialist

ethics. And indeed, Sartre, in considering the possibility that his student

could resolve his quandary by consulting a priest, makes just this criticism

of moral expertise. Thus he observes that the student, were he to seek

guidance from a priest, would first have to choose his advisor, and this

decision is both as much up to him and as determinate of what he will do as

his deciding the matter directly. He cannot, in other words, avoid having to

exercise his freedom by turning the decision over to someone else. For as

Sartre points out, different priests would give different advice, and inas-

much as the student would choose an advisor in view of that person�s faith,

principles, temperament, and sensibilities, he would decide what kind of

advice he would get. Choosing someone to be one�s advisor on such a

weighty personal matter is itself a reflection of one�s values and principles


and is therefore no less radical a choice than the choice of what action

to take.

The same points, existentialists should argue, apply mutatis mutandis to

using an ethical theory to determine what one should do. For different

theories give you different answers to this question, depending on the

circumstances you face, and inasmuch as in deciding to consult one theory

rather than another you choose a system of thought in view of the values

and principles it elaborates, you decide what kind of answer you will get.

Because each theory, as we have already noted in the several we have

examined, implies a different moral ideal, a decision to consult one theory

rather than another is a choice of an ideal of human life by which to live.

Choosing this ideal implies an aspiration to be a person of a certain sort. So

it, too, represents a radical choice. And to think otherwise, to think that

your decision to use a particular theory is correct because it squareswith the

facts of your situation or brings to bear on thematter at hand standards that

no sound deliberation on thatmatter can ignore, is to deny your freedom by

supposing that the matter is to be decided by inescapable facts and stand-

ards. From the viewpoint of existentialism, it is to choose in bad faith. This is

the criticism of ethical theories that is proper to existentialist ethics.

Sartre�s criticism, by treating the example of the student as showing that

ethical theories are necessarily too weak to provide useful guidance in

complex concrete situations, goes too far. In making it, he misreads the

import of his own example.

What of his further point that a radical choice does not follow from the

chooser�s seeing the values that he should pursue but rather acting on a

radical choice creates the values that define his pursuit? Here, too, Sartre

appears to havemisread the import of his example. This is evident from the

very fact of the student�s being in an ethical quandary. He is torn between

fighting for the Free French Forces and looking after his mother. Yet he

would not be torn in this way if he didn�t see value in both options. His

quandary consists in his being unable to decide between two courses of

action of whose values he has no doubt, patriotism and filial devotion. It

does not consist in his being unable to decide on a course of action that will,

as a result of his choosing it, have value. That condition would be better

captured in an example of someone in despair about life, someonewho sees

no value in anything he might do in life. Such a person might well be

advised to choose an activity in which he can immerse himself, even if he


sees no value in it, as a way of eventually coming to see value in things. But

such advice would be offered as therapy for someone who had fallen into a

funk and needs a means to escape it. It is not the sort of direction one gives

to someone who, like Sartre�s student, faces a situation in which he sees

value in each of two or more actions open to him and is conflicted about

which to pursue. The latter situation is a typical occasion for exercising

personal autonomy. Yet Sartre, it would seem, thinks that one�s integrity is

compromised if one chooses a course of action in view of the value one sees

in it, and this is surely a mistake. For a person who sees value in a course of

action is not compelled to pursue it. He is still free to turn away. In choosing

to pursue it, he affirms or embraces the value. He does not create it.

Some existentialist writers, notably Albert Camus (1913–60), have in

effect proposed that actingwith integrity requires taking up the perspective

of despair about life and choosing from it the values and principles bywhich

one will live. Only by first regarding the world as valueless and absurd and

then committing oneself to action from such a perspective does one take

full responsibility for one�s choices and actions, so these writers maintain,

and anything less than taking full responsibility for one�s choices and

actions is a loss of integrity. But this thought, too, is a philosophical extrav-

agance, similar, I suspect, to Descartes�s conceit that to attain knowledge he

had to start from a position in which he had no beliefs and find a corner-

stone from which to build a solid body of them. Even the person in despair

about life, although he cannot see value in anything, still has beliefs about

values. He knows, for instance, the difference between a boxing match and

a barroom brawl, or between the performance of a symphony and the

clamoring of taxi horns on a crowded New York City street. He can recog-

nize the event in the former of each of these pairs as a human activity

organized to exhibit something of value when it is done well, athleticism in

the one case, music in the other. And thus, he understands, if only tacitly,

how cultures, traditions, customs, and the like are the substrata of different

values in human life. So if Camus is proposing that integrity requires our

making radical choices from the perspective of despair about life, he is not

proposing that we suspend all belief about value in making these choices.

But in that case it is a mystery as to why he thinks choosing from such a

perspective is necessary for choosing with integrity.

Alternatively, he is proposing that integrity requires our making radical

choices from a perspective in which, not only does nothing appear to us to


have value, but all of our beliefs about value are suspended. That is, wemust

suppose that in making these choices not only do we not see any value in

the different thingswe could do ormake butwe also do not believe there is a

difference between human action and production that can be donewell and

the movements and products of random or purposeless human behavior.

Not only do we not see any value in music that we might produce, say, or

gymnastics that we might engage in, but we do not believe there is a differ-

ence between music and mere noise or gymnastics and the movements of

an epileptic during a seizure. On this proposal, it is by virtue of a radical

choice to do something ormake something thatwhat one does ormakes has

value. One invests it with value, so to speak, by the sheer act of one�s will in

choosing to do or make it. Thus, if a person chooses to devote his life to

driving a car down busy New York City streets and repeatedly honking its

horn, he gives value to the activity and to the noise he thereby creates. In

this case, then, there is no mystery as to why Camus thinks choosing from

such a perspective is necessary for choosing with integrity. For he must

think that choosing and acting with integrity means taking responsibility

not only for the action one chooses but also for its having value, and one

could not be responsible for its having value unless the valuewere a product

of one�s will. But if this is Camus�s proposal, then it is false if not incoherent.

Someone who chooses to play the violin, for example, can decide either

to play it well, if he knows how, and produce music, or to play it badly and

produce mere noise. But if he does the latter, he cannot make it the case, no

matter his intentions, that he is playing the violin well. He cannot make it

the case that themere noise hemakes has value. Though what he produces,

music or noise, is up to him, whether or not it has value is not. Similarly,

while he can choose to drive a car down a busy New York City street and

repeatedly honk its horn, his doing so does not give the sounds he creates

value. Honking a car horn for the purpose of warning pedestrians or other

drivers is something one can dowell or badly according as it is done in away

that augurs success or failure. The sounds one makes in that case can have

value as ameans of warning others. But honking the horn for its own sake is

not something that can be done well or badly, and correspondingly the

sound one makes in this case cannot have value in and of itself. To think

otherwise, to think that one can give such horn honking value simply by

choosing to do it, is somewhat like Humpty Dumpty�s belief, in Through the

Looking Glass, that the words he speaks can have whatever meaning he


chooses to give them simply by virtue of his so choosing.20 The impossibility

of this belief is transparent. While one can choose what words one speaks,

one cannot choose what they mean. Camus�s proposal, on this second

construction of it, is similarly impossible.

10. Existentialist ethics pruned of excess

What, then, is left of existentialist ethics once one removes its excesses? At

its core there is the doctrine of radical choice. This doctrine, when under-

stood apart from the faulty lessons about ethical theories and the origin of

values that existentialists tend to draw from it, is the view that certain

decisions, decisions about what sort of person to be and by what values

and principles to lead one�s life, cannot bewrong if the thinking fromwhich

they follow is clear and cogent, if they are based on a correct understanding

of the facts of one�s situation, and if one is sincere in theway oneweighs the

different considerations that bear on them. This is a not an implausible

view. Nor is it trivial. For it contradicts the belief that some ends in life and

some principles of right conduct are not optional if one is a rational agent.

To decline to pursue the former or comply with the latter, according to the

philosophers who hold this belief, goes against reason. One thing that

remains, then, is a powerful thesis about the limited scope of practical

reason. On this thesis, to exercise practical reason is to determine means

to some end and not to determine an end whose achievement is not also a

means to some further end. Or as Aristotle put it, deliberation is of

means only.21

At the same time, existentialist ethics assumes that one�s ultimate ends

are themselves an object of choice. It thus rejects the conception of human

action Aristotle assumed in asserting his thesis. For in asserting it, Aristotle

meant to be observing that people do not choose the ultimate ends of their

actions. To the contrary, he held that every human action has the agent�s

well-being as its ultimate end.22 Hence, choice must always be choice of a

means to this end. This view plainly contradicts the doctrine of radical

choice. Existentialist ethics, in attributing personal autonomy to human

beings, takes determination of the ultimate ends of one�s actions to be a

20 Lewis Carroll, Through the Looking Glass (and What Alice Found There), ch. 6.21 Aristotle, Nicomachean Ethics, 1112b11–12. 22 Ibid. 1095a13–20.


matter for the agent to decide. The ultimate ends of one�s actions are thus

optional in the sense of its being open to a rational agent, consistent with

his being rational, to choose what his final ends are. In other words, there

are no ends he must pursue as a condition of his being rational.

Existentialist ethics, then, once one removes its excesses, consists of two

theses. One is a thesis about personal autonomy: that there is no end (such

as one�s own well-being) whose pursuit as an ultimate end in one�s life is a

condition of one�s being rational. The other is the thesis about practical

reason�s limited scope: exercises of practical reason are limited to determi-

nations of means to one�s ends. This pair of theses, even though lacking the

audacity of the views from which it descends, constitutes nonetheless a

significant viewwithin contemporary ethics. Other views, corresponding to

the teleological and deontological theories of ethics that moral philoso-

phers have expounded, take human rationality or practical reason as fixing

the ultimate ends men and women ought to pursue in life or determining

the principles they ought to follow in pursuit of the ends aroundwhich they

have organized their lives. Each of these theories yields amoral ideal whose

realization gives meaning to their pursuing those ends and complying with

those principles. Existentialist ethics, by contrast, treats all such moral

ideals as objects of choice similar in this respect to other ideals of human

existence that a person might choose as a model of how to live. Neither

human rationality nor practical reason fixes or dictates which is to be

chosen. The choice rather is a radical one. Ethical theories then become

guides to realizing a moral ideal, and one comes to accept an ethical theory

as a result of having embraced the moral ideal it yields rather than having

decided that it represents the soundest account of morality. Existentialist

ethics can thus acknowledge a role for ethical theory in an autonomous life,

though it is not the traditional role of systematizingmorality and supplying

the grounds on which its standards are rationally justified. The challenge it

represents to traditional ethics is a challenge to the very idea of morality as

the subject of a philosophical study whose aim is to formulate its basic

standards and to determine the rational grounds of their authority to

govern our lives.


7 Practical reason

1. Meta-ethics

Our study of different ethical theories began with reflection on a hypo-

thetical question about whether you would have good reason to do the

honest thing if you found a lost purse containing a huge wad of cash. The

question generated an inquiry into the rational basis of the standards of

honesty and justice. These standards, and standards of morality generally,

appear to have an authority in our lives superior to that of the conventional

standards embedded in local social practice. For one thing, they appear to

have such authority because we measure our local social practices and the

conventional standards embedded in them against standards of justice and

morality. For another, we commonly praise and honor those who resist

conforming to conventional standards when those standards create privi-

leges for some and hardships for others that are unjustly enjoyed and borne.

This is especially true where such resistance carries personal costs like

ridicule, ostracism, or worse. Huck Finn�s actions in helping Jim in his

attempt to escape slavery are exemplary. The assumption of our inquiry,

then, has been that the authority of the standards thatmorality comprises is

founded on reason and truth and not on mere custom or prejudice.

Otherwise it would be hard to explain its superiority. Accordingly, we

have treated the different ethical theories we surveyed, up to existentialist

ethics, as systems of thought constructed to confirm this assumption. As

such, they represent competing accounts of morality as comprising stand-

ards of conduct and character whose authority has rational foundations.

Existentialist ethics is different, however. It does not offer such an

account. To the contrary, its doctrine of radical choice opposes the assump-

tion of the inquiry in which we have been engaged and which has produced

such accounts. It is not hard to see why. The doctrine derives from a con-

ception of individual self-determination according to which each of us, as

196

individuals, is the final authority on questions about what sort of person to

be and by what values and principles to guide our lives. On this conception

there are no criteria by which to assess someone�s answers to these ques-

tions as right or wrong independent of the sincerity of those answers and

the soundness of the deliberations that yield them. Each person alone

determines the actions it is imperative for him to do. Were he to look

instead to external standards in deciding what to do, standards he regarded

as authoritative independently of his own decision to follow them, he

would be denying the very freedom he has to decide how to live. He

would be denying his own autonomy. Implicitly, then, existentialist ethics

denies that morality comprises standards of conduct and character whose

authority is founded on reason and truth. Rather it holds that such stand-

ards have authority in a person�s life only by virtue of the person�s sincerely

deciding to live his life by them. Consequently, according to existentialism,

inquiries in ethics into the rational basis of the authority these standards

have, inquiries of the kind with which our study began, are essentially

mistaken.

The opposition of the doctrine of radical choice to the assumption on

which traditional inquiries in ethics proceed presents us with questions of a

new kind. Does our capacity for personal autonomy imply that no answer to

a question about what it is right to do or what one ought to do is ever true

independently of our attitudes towards the different actions we can do? Are

disagreements about whether an action is right or ought to be done resolv-

able, at least in principle, through consideration of the relevant facts and

what they imply? And is universal knowledge of right and wrong possible?

What makes questions like these different from questions of right and

wrong that arise in the course of everyday life and that lead, whenwe reflect

on them, into ethical inquiry is that they turn the very project of ethical

inquiry into an object of study. Accordingly, they introduce a higher level of

inquiry. Each asks something about the very nature of the everyday ques-

tions of right and wrong – what I will refer to as the ground-level questions of

ethics – with which traditional ethics is concerned. Thus the first asks, in

effect, whether these ground-level questions of ethics admit of objectively

true answers given human freedom. The second asks whether there is a

rational procedure or method by which disputes over such questions can,

given enough time and factual knowledge, be resolved. And the third asks

whether answers to them could ever qualify as human knowledge. Other

Practical reason 197

higher-level questions onemight ask are whether ground-level questions of

ethics are essentially different from questions of natural science and

whether one�s answers to them, if sincere, entail some disposition of

one�s will. These and similar higher-level questions define the field of

meta-ethics.

Ground-level questions of ethics include not only questions about right

and wrong but also questions about values. We want to know not only

whether one ought always to do the honest thing in circumstances in

which dishonesty would be more profitable but also whether anything

has value in itself besides pleasure and freedom from pain, whether, for

instance, human life would still be of infinite value, as we say, if it excluded

experiences of pleasure and pain. Such questions about values are also the

objects of meta-ethical inquiry. Typically those engaged in such inquiry do

not distinguish between these two kinds of ground-level question.

Typically, that is, they ask the same higher-level questions about both

kinds. Nonetheless, it is useful to distinguish between them, for the consid-

erations that bear on how the higher-level questions are answered may

differ according to whether one asks them about ground-level questions

concerning right andwrong or ground-level questions concerning values. In

particular, while the doctrine of radical choice bears on how one answers

higher-level questions about the former, it has no bearing on how one

answers the same questions when asked about the latter. It has no bearing

because the freedom the doctrine affirms, while it entails the freedom to

decide whether to pursue some values and ignore others, does not entail

freedom to decide what has value and what does not. While it calls the

authority of moral standards into question, it yields no such consequence

for the reality of values. Failure to appreciate this, as we saw, led existenti-

alist thinkers like Camus grossly to overstate the implications of their ideas

about human freedom.

Accordingly, we need to limit our examination of the challenge to tradi-

tional ethical inquiry that the doctrine presents tometa-ethical questions as

they concern ground-level questions about right and wrong. Let us call the

answers to these ground-level questions normative statements or directives.

They are to be distinguished, then, from the answers to ground-level ques-

tions about values, which I�ll call evaluative statements or valuations. It will also

be useful to expand the class of normative statements to include directives

that are not answers to questions about right and wrong, that is, what


morality requires and prohibits, but rather are answers to questions about

personal and practical matters. Accordingly, the class of normative state-

ments includes all hypothetical imperatives, in Kant�s sense, as well as any

that are categorical. It includes as well ordinary pieces of advice such as

�You should buy the tickets now; the show will be sold out soon� and �You

ought to keep a flashlight in the car in case of emergencies.�

The advantage of expanding the class of normative statements to include

all hypothetical imperatives and other pieces of practical advice that do not

concernmoralmatters is that the conditions under which such directives are

valid imperatives or sound bits of advice are evident. They therefore repre-

sent either a good model for understanding the conditions under which

directives that concern moral matters are true or a good contrast class – as

Kant saw – for the samepurpose. In general, normative statements aremeant

to guide the choices and actions of the people to whom they are addressed. If

the statements are true, then they offer those to whom they are addressed

reason to choose to act as they direct. In other words, their being true implies

that one could come to this choice on the basis of sound deliberation. Hence,

we can formulate the challenge that the doctrine of radical choice presents to

traditional ethical inquiry as a challenge to the thesis that some choices and

actions are the outcome of deliberations that would be sound nomatter who

the deliberator was or what his or her personal feelings or attitudes were. In

short, it is a challenge to the idea of universal truths of practical reason, true

normative statements that apply to every rational agent.

2. Meta-ethical disputes: an illustration

One answer to this challenge is to deny that human beings have freedom of

the kind that the doctrine of radical choice implies. If you accept psycho-

logical egoism as Hobbes did, and hold that all voluntary actions are self-

interested, that the aim of every voluntary action a person takes is to

advance his interests,1 then you deny that people have this freedom. For

in agreeing with Hobbes, you are agreeing that no one can voluntarily act

contrary to his or her own interests, and if people have the kind of freedom

the doctrine of radical choice implies, then they can deliberately choose to

do something that hurts their interests and to act on that choice.

1 Hobbes, Leviathan, ch. 14, par. 8.


This implication of accepting psychological egoism carries over, then, to

the opposition between Hobbes�s thesis that the laws of nature are true

directives applying to all rational agents and the existentialists� skepticism

about there being such directives. For if you accept psychological egoism,

then you believe with Hobbes, at least implicitly, that there are such direc-

tives. The reason is clear, as Hobbes explained. People have as their primary

interests the necessities of life: food, shelter, clothing, mobility, safety, and

the like, and in some situations, some actions will be more conducive to

securing those interests than others. Accordingly, directives to do the for-

mer rather than the latter in those situations will be true and apply to all

rational agents. The brilliance of Hobbes�s ethics lies, then, in the argu-

ments hemade to show that the basic standards of right and wrong, such as

those of honesty, gratitude, and fairness, qualify as directives of this sort.

Their brilliance notwithstanding, however, the arguments have no force if

one accepts the doctrine of radical choice. Since, according to this doctrine,

a rational agent need not take his own interests as an end to be pursued, the

basic standards of right and wrong need not apply to every rational agent.

The lesson in this clash of meta-ethical views is twofold. First, Hobbes�s

program of deriving basic standards of right andwrong from the fundamen-

tal principle of egoism is a good example of a program in ethics that takes

sound pieces of personal advice as amodel for understanding the conditions

under which directives that concernmatters of right andwrong are true. On

this model, if you know a person�s aims and do not think they are contrary

to his interests, then sound advice consists in a directive to do what is

necessary or most conducive to his realizing those aims. If a friend, say,

who wants to visit a foreign country and knows nothing about traveling

abroad comes to you for advice, one piece of sound advice you could give

him is that he ought to get a passport. This would be sound advice, a true

directive, since his aim is to visit a foreign country and having a passport is

necessary for achieving that aim safely.2 Hobbes�s program, in effect, gen-

eralizes on this model. It starts with very general aims that Hobbes attributes

to all human beings – those of securing for oneself the necessities of life – and

then offers arguments to show that following fundamental standards of right

2 On Kant�s account of hypothetical imperatives, �You ought to get a passport� is a valid

hypothetical imperative for someone whose end is to visit legally a foreign country, for

having a passport is a necessary means to legal travel to another country.


and wrong is necessary for achieving these aims. Fittingly, then, in Hobbes�s

terms, these standards are to be taken as counsels rather than commands.3

Second, the opposition between Hobbes�s program and the existen-

tialists� skepticism about true directives applying to all rational agents is

an instance of a general opposition between such skepticism and teleolog-

ical theories of ethics that are premised on doctrines, like psychological

egoism, that restrict the motives on which humans act to those of a certain

kind or restrict the ends that humans pursue to those of a certain character.

All teleological theories of this kind explain the truth of directives about

matters of right and wrong on the model of sound advice. All take such

directives as hypothetical imperatives whose validity rests on the attribu-

tion of certain motives or ends to every man or woman and in the correla-

tive thesis that these at bottom are the only motives or ends in the agent�s

psychology. Each then answers the challenge that the doctrine of radical

choice presents to traditional ethical inquiry by denying that human beings

have freedom of the kind that the doctrine implies. The most influential of

these teleological theories is Aristotle�s eudaimonism. Among such theo-

ries, it offers, perhaps, the strongest opposition to the doctrine, and for this

reason its answer to the doctrine merits careful study.

3. Aristotle�s answer and an existentialist response

Aristotle held that the final end of every action is the actor�s well-being. No

one, on his view, can aim to make his own life worse or to bring harm

overall upon himself. Hence, it is not, on this view, open to people to choose

a life of misery or to seek their own ruin. It is no more open to them to

choose such a life or to seek such an outcome than it is open to them to

choose immortality or to seek to live the life of a blue jay. Of course, the

limits to what humans can choose or seek in these latter cases are due to our

nature as physical and biological beings, and no adherent to the doctrine of

radical choice disputes the existence of such limits. In the former cases, by

contrast, the limits, if they exist, are due to our nature as rational beings,

and adherents to the doctrine dispute any view that takes human rationality

to be the source of such limits. Aristotle�s view, in other words, yields limits

on human freedom that are contrary to what the doctrine implies.

3 See p. 128 above and Hobbes, Leviathan, ch. 15, par. 41.


Few contemporary philosophers, however, accept Aristotle�s view. Most

allow, as at least tenable in an ethical theory, that it represents morality as

requiring people in some circumstances to sacrifice their ownwell-being for

the greater good or for the cause of justice, and such a theory would be

untenable if it were impossible for people intentionally to act against their

own good.4 In addition, studies in literature, psychiatry, and clinical psy-

chology of men and women driven by self-destructive or self-belittling

motives have persuaded most contemporary philosophers to abandon the

view of human beings as always and necessarily striving to better their lives

and promote their own good. Indeed, one such study, Fyodor Dostoyevsky�s

novella Notes from Underground, is among the most important sources of

existentialist thought. Its protagonist, a former lowly official in the Russian

civil service, recounts, as if writing in a diary, the many times he deliberately

acted contrary to his best interests. He relishes, it seems, telling of how he

often acted out of spite to get revenge on those who had annoyed him and of

his tendency to court humiliation and suffering. �Sometimes,� he declares,

�a man is intensely, even passionately attached to suffering – this is a fact.

About this there is noneed to consult universal history: ask yourself if you are

a man and have ever lived even in some degree. As for my own personal

opinion, I find it somehow unseemly to love only well-being. Whether it�s a

good thing or a bad thing, smashing things is also sometimes very pleasant.�5

This compelling portrait of a small-minded and mean-spirited contrarian, a

4 Youmight think Aristotle�s ethics, being a version of Platonic eudaimonism, is not open

to this objection. After all, on any version of Platonic eudaimonism, doingwhatmorality

requires and acting to advance one�s well-being cannot come apart. But because

Aristotle, unlike the Stoics, say, held that virtue does not guarantee that onewill achieve

the highest good, because he allows that a virtuous person is still vulnerable to tragic or

grievous loss in life, there is still a question of whether, on his ethics, the two can come

apart. Hence, tomaintain, as he does, that all actions aim at the agent�swell-being and so

to keep them from coming apart, he must hold that to act shamefully always damages

one�s well-beingmore than suffering a great personal loss. And tomaintain this thesis, it

would seem, is too high a price to pay for keeping them from coming apart. Surely not

every shameful act does worse damage to one�s well-being than the most grievous and

tragic of losses. If Agamemnon is required by morality to sacrifice his daughter to the

gods, the loss he suffers from killing his own daughter may surely be a greater blow to

his well-being than the disgrace he would have suffered if he had declined to make the

sacrifice.5 Fyodor Dostoyevsky, Notes from Underground/The Double, J. Coulson, trans.

(Harmondsworth, England: Penguin Books, 1972), p. 41.


not uncommon product, Dostoyevsky (1821–81) assures us, of nineteenth-

century urban European life, confounds any view of human beings,

Aristotle�s included, as constant seekers of their own good and vigilant

guardians of their well-being.

While Aristotle�s view of human action as in every case having the

agent�s well-being as its final end now has few adherents, there is a weaker

view that he also holds and that could serve, too, as a basis for denying that

men and women have the kind of freedom the doctrine of radical choice

implies. This weaker view enjoys much greater support among philoso-

phers. Some may even see it as a piece of common sense. This, at any rate,

is how Aristotle saw it. Thus book I of his Nicomachean Ethics opens with the

statement that every action and every pursuit aims at some good.6 Aristotle

immediately and problematically infers from this statement that there is

some good at which all actions and pursuits aim, and this inference is what

leads him to the stronger view that the aim of every human action and

pursuit is the actor�s own well-being. But one need not follow Aristotle in

making this inference in order to find in his account of human action

ammunition with which one can attack the doctrine of radical choice. The

statement with which book I of the Nicomachean Ethics opens alone provides

it. For it entails that a person cannot choose as a final end of action a

condition or activity that he regards as either without value or positively

bad. It places, in other words, limits on human freedom of a kind that is

contrary to what the doctrine of radical choice implies.

Nor should there be any doubt that those who advance this doctrine

would see the opening statement of book I of the Nicomachean Ethics as

antithetical to their understanding of our capacity for personal autonomy.

They would see it, that is, as expressing a view someone might hold

unreflectively because the idea of a person�s choosing to be an enemy of

what is good is beyond the grasp of conventional thought. Or perhaps they

would see it as expressing a view someone might hold defensively as a way

to hide from himself possibilities too unsettling to be faced. In any case,

they would maintain that the question of whether to be a friend or an

enemy of what is good or merely a neutral party is one that a person can

and should decide for himself or herself. The capacity for personal

autonomy, though it does not, as I argued in the last chapter, include the

6 Aristotle, Nicomachean Ethics, 1094a1–3.


capacity to determine what is good, includes at least this much. And the

person who thinks he is bound to pursue only that which he sees as good is

not yet, they would argue, fully exercising this capacity. He is not yet a self-

determining agent.

Camus, for instance, sees the capacity in this way. Recall that the

Archimedean point of his philosophy is the perspective from which the

world appears absurd. It is the perspective of existential despair. Why,

Camus wonders, once one has taken up this perspective, should one con-

tinue with one�s life? This question, he thinks, the question of whether to

commit suicide, is the first question of practical philosophy. When nothing

in theworld appears tomatter, onemustfirstfind an answer to the question

why go on.7 And Camus treats the question as truly open. That is, he takes

suicide to be an option a man who sees the world as absurd might actually

choose, not because the man thinks he could achieve some good by ending

his life, but because all of the other options available to him appear equally

pointless. Hence, if the man were to choose to commit suicide, his ending

his life would be an action whose end he did not see as good. It would

therefore confound Aristotle�s thesis that the end of every action or pursuit

is something the agent sees as good. For Camus, then, the first question of

practical philosophy defines, at least for those who confront it, a radical

choice, a choice that Aristotle�s ethics cannot comprehend.

Of course, the idea that suicide is something a person could deliberately

commit on account of his seeing nothing worth doing in life may just be an

existentialist conceit. It may represent nothing more than an abstract pos-

sibility. After all, it is one thing to have the idea of doing a certain action,

and quite another to be able to bring oneself to do it. For one�s being able to

do an action in circumstances of the kind that call forth sufficient motive

for doing it does not mean that one can do it in any circumstance in which

there is nothing physically stopping one from doing it. Think, for instance,

of deeply religious people who martyr themselves in protest like the

Buddhist monks in Vietnam during the Diem regime who immolated them-

selves to protest the regime�s persecution of Buddhists. That they could set

fire to themselves in such circumstances does not mean that any of them

could have brought himself to do the same thing in peaceful circumstances

7 Albert Camus, �The Myth of Sisyphus,� in The Myth of Sisyphus and Other Essays, J. O�Brien,

trans. (New York: Vintage Books, 1955), p. 3.


when Buddhists had no complaints against their government. Nor, needless

to say, does it mean that anyone who gets the idea of self-immolation from

their martyrdom could walk out onto a public plaza, douse himself with

gasoline, and set himself on fire. So too that there are circumstances that

drive people to commit suicide privately does not mean that existential

despair is one of them. Camus treats suicide as if it were a real choice that

someonewho came to an understanding and view of theworld as a hopeless

arena of meaningless phenomena might make, but his doing so may just

reflect his overestimating the agential power of human beings to realize the

ideas of actions they can contemplate doing. We should not therefore take

Camus�s treatment of suicide as a refutation of Aristotle�s thesis that all

actions aim at some good.What it does show is that rejection of the thesis is

basic to his philosophy and that of like-minded existentialists.

4. Can there be motives that aim at doing evilfor its own sake?

How then shall we resolve the opposition between the thesis withwhich the

Nicomachean Ethics begins and the existentialists� doctrine of radical choice?

Fortunately, its resolution may not require our deciding the question about

human agential power that Camus�s treatment of suicide raises. For that

question is equivalent to the question of whether human action can ever be

motiveless, and the thesis with which the Nicomachean Ethics begins, the

thesis that every human action aims at some good, is stronger than the

thesis that every human action springs from some motive. Hence, we can

examine the soundness of the thesis by considering the possibility of a

motive to act that does not imply that the actor is aiming at some good. If

such a motive is possible, then the thesis is unsound, and we will therefore

have removed the objection to the doctrine of radical choice that it yields

without having to take up the question about human agential power that

Camus�s treatment of suicide raises.

For this purpose, let us consider a version of the thesis due to Thomas

Aquinas (1224/5–74): whatever is desired is desired sub ratione boni, that is,

under an aspect of goodness.8 This version nicely reveals a source of equiv-

ocation in Aristotle�s view. On the one hand, one may desire an object

8 Aquinas, Summa Theologica I–II, q. 8, a. 1.


because one sees something good in it. That is, it is the object of one�s desire

because there is something good in it that one sees and one�s seeing it

excites the desire. On the other hand, one may see something good in an

object because one desires it. That is, when one desires something, onefinds

the thought of one�s having it in whatever way would satisfy that desire

agreeable or pleasant, and if the thought of having something is agreeable

or pleasant, then one is disposed to see that thing as good. Thus, in this case,

too, the object is desired under an aspect of goodness. Since the end of an

action is the attainment of the object of whatever desire is themotive of that

action, Aristotle�s thesis is equally equivocal. It may be true, on the one

hand, because we act only from desires whose objects we desire because we

see something good in them, or it may be true, on the other, because the

desires from which we act provide the ends of our actions and we see

something good in those ends because we desire the objects that constitute

them.

The opposition between Aristotle�s thesis and the doctrine of radical

choice depends on which of these ways, if either, the thesis is true. The

thesis stands in substantial opposition to the doctrine if it is true in virtue of

our seeing something good in the objects of our desires independently of

our desiring those objects. For in that case, because we are limited in what

we can desire towhatwe can see to be good independently of our desiring it,

we are limited, in what we can choose to do, to actions whose ends we see,

independently of that choice, to be good in some way. And while we might

pursue ends that are in fact bad, we can do so only under themisperception

of them as good and thus only in error. This, then, is a substantive limit on

our freedom of the sort whose existence the doctrine of radical choice

denies.

If, on the other hand, one takes Aristotle�s thesis to be true on the

grounds that we regard anything we desire as good in virtue of our desiring

it, then the thesis does not stand in any real opposition to the doctrine. The

thesis, when taken in this way, places no limits on what we can desire, so it

does not limit us, in what we can choose to do, to actions whose ends we

must perceive independently to have some choice-worthy property. In

being free to choose whatever we desire and to desire whatever appeals to

us regardless of the properties it has independently of that appeal is as

much freedom as the doctrine of radical choice need imply. To be sure, as

we�ve seen, the doctrine may also imply freedom to choose to act


independently of one�s desires, which is to say, independently of all

motives. But the possibility of such choice is a highly contentious meta-

physical thesis, and for that reason, the opposition between the doctrine

and Aristotle�s thesis that it creates may be treated separately. Leaving it

aside, then, we may conclude that Aristotle�s thesis opposes the doctrine

only if it is true by virtue of our seeing something good in the objects of our

desires independently of our desiring those objects.

Aristotle�s thesis, however, if it is taken to be true in this way, is open to

the same kind of objection that throws his stronger view that all actions aim

at promoting the agent�s well-being into doubt. Men who have soured on

life, like Dostoyevsky�s underground man, not only exemplify people

inclined to actions that do not aim at promoting their own well-being but

exemplify as well people inclined to actions whose final end is to spoil or

damage something they see as good. And because the very point of these

actions is to spoil or damage something good, they go against Aristotle�s

thesis that every action or pursuit aims at some good. When, for instance, a

person�s youthful aspirations and ambitions for himself go bad, when

things don�t turn out for him as he once expected, which is to say dreamed,

the disappointments can come hard and embitter. Such bitterness about

one�s reduced prospects in life typically leads to petty and nasty conduct

toward others and to conduct that is self-abusive and self-injurious. The

underground man is a case in point. Recall his earlier remark in which he

enthusiastically comments on howhe likes smashing things evenwhen this

is a bad thing to do. He has lost interest in his life and esteem for himself,

and his self-contempt manifests itself in degrading and destructive actions.

These are not actions whose ends the underground man sees as good.

Dostoyevsky asks us to look beneath our social selves, the personae, as it

were, that we inhabit and to some extent shape in engaging publicly with

others. Our actions in such engagements are generally construed as either

competitive and self-advancing or cooperative and constructive. In either

case, they easily fit Aristotle�s thesis. We readily understand socially struc-

tured actions as having as ends the purposes embedded in those structures,

and those purposes are beneficial either to the individual agents or to

society generally. Buying a pack of gumor a candy bar at a local convenience

store is a mundane example of the former. Discarding gum or candy wrap-

pers in public trash cans is an example of the latter. Our lives are filled with

such actions, and even when we are alone, in the privacy of our homes, the


conventions of housekeeping and of the preparations we make to present

ourselves to others define much of what we do. But not everything we do

contributes to our public engagements with others. Privacy is not only

something we require to prepare for public engagements but is also a

cloak for actions that make no contribution to our self-presentations.

These are both actions that take place in private spaces and are done for

their own sake – actions such as making faces at oneself in the mirror or

sharpening all the pencils in a desk drawer, even though one has no need of

sharp pencils – and actions done in public space whose ultimate ends are

disguised and whose true motives are kept hidden – actions such as a

malicious prank one secretly plays on a co-worker who has rebuffed one�s

advances or an accident one who cannot cope with success provokes to

insure failure. If we think of our public lives as conducted openly and above

ground, then these are actions that we must unearth and bring to light to

observe. They are actions that generally need a cover of darkness to be done;

if done publicly, they would be a source of embarrassment or shame. It is

surely with this point in mind that Dostoyevsky characterizes the protago-

nist of his strange novella as an underground man.

Aristotle, who conceived of men and women as social animals, likely

came to his views of human action from this conception. While he does not

require a social context for the intelligibility of actions, it is plausible to read

him as regarding actions as always fitting within some social structure. One

bit of evidence for this reading is his assuming without argument that

justice and honesty are moral virtues, which is to say, excellences of

rational functioning. That Aristotle seems untroubled by the core problem

of Plato�s Republic, the problem Glaucon and Adeimantus set for Socrates,

strongly suggests a view of rational action that cannot accommodate the

anti-social strain in Dostoyevsky�s underground man. It suggests a view of

deviations from the norms of human social life, like those the underground

man at once exhibits and applauds, as symptomatic of deficiencies in

practical reason rather than the heterogeneity of human motivation.

Perhaps, then, Aristotle�s thesis that all actions and pursuits have some

good as their end reflects preconceptions that preclude Aristotle from see-

ing those motives and actions that do not fit his view. Perhaps, they blind

him to certain kinds of motives and actions in human life that are typically

hidden from public view. Alternatively, though, his thesis may simply

reflect the fact that he never knew any Russians.


5. The obsolescence of Aristotle�s answer

Counterexamples have only so much force in philosophical argument.

Dostoyevsky�s underground man is no exception. Confronted with his

actions and similar examples, diehard Aristotelians will respond by inter-

preting themotives of those actions consistently with Aristotle�s thesis. The

underground man, by his own admission, takes pleasure in his suffering

and humiliation, so, they will argue, he makes himself suffer or seeks

humiliation for the sake of this pleasure. His actions thus aim at some

good. Or again, the man who plays a malicious prank on a co-worker who

has rebuffed his advances does so, they will argue, for the sake of exercising

power over his co-worker, and having and exercising power over another is

typically seen as a good. These interpretations are plainly necessary to

maintain Aristotle�s view, and some such interpretation will always be

available if only because it is reasonable to suppose, as we noted, that one

generally takes pleasure in attaining the objects of one�s desires. Hence, any

action on a desire is interpretable as an action in which one seeks pleasure

through attainment of the object of that desire.

Nonetheless, we have good reason to resist these Aristotelian interpreta-

tions. To have merit, they must be more than reflexive applications of a

dogmatic belief that every human action aims at some good. Specifically,

they must rest on some deeper account of why every human action essen-

tially has such an end. Yet there is no such account that is consistent with

understanding human actions as caused by the motives from which they

spring. Motives such as love and compassion are, of course, naturally under-

stood as prompting actions that aim at some good, for each implies a desire

to benefit its object. Love implies a desire to make one�s beloved happy.

Compassion implies a desire to relieve the suffering of the individual with

whom one commiserates. But motives such as hatred, spite, envy, and

malice are not readily understood as prompting actions that aim at some

good. To the contrary, because each implies a desire to hurt its object, we

naturally understand them as prompting actions that aim at what is bad.

Nor is it necessary for so understanding them that one suppose that the

person who acts under their influence sees something good in hurting

others or in making their lives worse. When we understand actions as

caused by the motives from which they spring, we understand that the

desires the motives imply explain, either by themselves or in concert with


certain beliefs, those actions. For to have a desire to do something is to be

disposed to do it in circumstances that appear to one propitious for doing it,

and consequently what explains one�s doing it is this disposition together

with beliefs that the circumstances favor one�s doing it. One�s seeing it as

good or as accomplishing some good has no role in this explanation. The

desire and these beliefs alone suffice.

There is the temptation, as we have seen, especially when malicious

desires are cited as counterexamples to Aristotle�s thesis, to defend it by

identifying such desires as at bottom desires for the pleasure one obtains

from doing the action, from hurting others or making their lives worse. But

to defend the view in this way is to confuse two distinct desires. Recall

Butler�s criticism of psychological hedonism from chapter 2. Because the

pleasure one gets from hurting others presupposes the desire to hurt them,

a desire for that pleasuremust be a different desire. And the same point will

apply wherever one tries to deflect counterexamples to the thesis by rede-

fining desires to do certain actions or obtain certain external objects as

desires for the pleasure that doing the action or obtaining the objects will

bring. It may be true that whenever an action aims at pleasure it aims at

some good, but given that not every desire is a desire for pleasure and some

are desires formaking another suffer, the assertion that every action aims at

pleasure has nothing to recommend it.

Indeed, the general strategy of defending Aristotle�s thesis by redefining

any desire offered as a counterexample to it as a desire for some apparent

good misconstrues the problem Aristotelian interpretations of these coun-

terexamples face when we understand human actions as caused by the

motives from which they spring. For the problem is that when human

actions are so understood, one does not need to suppose that their agents

regard their ends as good to explain them. Citing the desire implicit in the

motive along with relevant beliefs about the circumstances of action is

sufficient. Knowing, for instance, that someone desires a certain object

and believes that he can attain it in these circumstances by taking certain

actions is all one needs to explain why the agent takes those actions.

Consequently, redefining the person�s desire so that it is understood as a

desire for some good adds nothing to the explanation. It becomes a strictly

ad hoc maneuver, undertaken solely for the purpose of saving Aristotle�s

thesis. To defend Aristotle�s thesis, then, requires opposing, as inadequate,

explanations of human actions that consist in citing the motives from


which the actions spring. Or what comes to the same thing, it requires

opposing, as problematic, the conception of human action behind such

explanations. The explanations are causal in the sense that events and

conditions preceding an action are cited to explain it, so the allegedly

problematic conception behind them is that of an event the cause of

which consists in prior events and conditions. To oppose this conception

onemust believe that human actions are events whose nature does not lend

itself to causal explanations. One must believe, in other words, that explan-

ations of actions must take a different form.

Unsurprisingly, the most common defense along these lines appeals to

explanations whose form is characteristic of Aristotelian naturalism. These

are teleological explanations. Such explanations presuppose that the action

being explained is done intentionally. The relevant conception of human

action, then, is that of an intentional action. One explains an intentional

action by citing its point, what the agentmeans to do or achieve in so acting.

The point of an action is not a prior event or condition. It is, rather, what

gives the behavior its meaning or intelligibility as a human action. Hence,

the form of explanation is different from that of causal explanation. In

giving such an explanation, one assumes that the agent saw something

desirable or worthwhile in acting as he did, and the feature that makes

the action seem desirable or worthwhile to the agent defines the action�s

point. For instance, when we explain a man�s opening a window by saying,

�He is letting in some fresh air,� we give the point of his opening the

window, and we recognize it as such because we understand that removing

a barrier to fresh air�s entering the roomwill make the roommore comfort-

able. Specifying what is or seems to the agent to be desirable or worthwhile

about the actionmakes sense of it as something onewould intentionally do.

Teleological explanations therefore uphold Aristotle�s view. The action

necessarily aims at some good in the sense that its end is necessarily some-

thing that appears to the agent to be desirable or worth achieving.

In teleological explanations the idea of an action�s having amotive that is

its cause is replaced by the idea that to act intentionally is to act under the

guidance of practical reason, and to be a rational agent, that is an agent

guided by practical reason, is to be a seeker or pursuer of what is good.

People�s appetites and emotions, which on a causal explanation of human

action are or are the source of an action�s motives, are on a teleological

explanation states in which rational agents perceive or sense something


good or bad. And while these perceptions are liable to error owing to the

primitive, violent, or erratic character of the states of appetite and emotion

behind them, neither they nor those states can change one�s nature as a

rational agent. They cannot turn one away from seeking or pursuingwhat is

good. Thus, hatred, spite, envy, and the other emotions that, on a causal

explanation of human action, can prompt actions aimed at doing some-

thing bad, are on a teleological explanation states that affect what a rational

agent perceives as good or bad but do not change the agent�s orientation

toward bringing about what is good. In short, a rational agent, on the

conception of rational agency on which teleological explanations are

based, essentially functions as a seeker or pursuer of what is good.

Accordingly, one explains an intentional action by pointing out how the

action fulfills this function.

Any teleological conception of nature is now obsolete. Consequently,

either teleological explanations of natural phenomena are replaceable by

causal explanations and used solely as proxies for the latter, or they are

anachronisms. A defense of Aristotle�s thesis that rejects causal explana-

tions of human action in favor of teleological ones therefore bears the

burden of justifying its rejection of causal explanations. It has the burden

of showing that the view is not anachronistic. Hence, the justification

cannot consist in wholesale rejection of causal explanations of natural

phenomena in favor of teleological ones, since that would be tantamount

to affirming a teleological conception of nature. It must therefore consist in

first distinguishing intentional action frommere behavior, whose status as

natural phenomena is indisputable and is thus explicable by the prior

events and conditions that are its causes, and then showing that what

distinguishes intentional action frommere behavior alsomakes intentional

action resistant to causal explanation. The distinction between the two that

defenders of Aristotle�s thesis would draw is plain. Intentional action has a

point, they would say, whereas mere behavior does not. So the burden they

bear is to show why intentional actions� having a point makes them resist-

ant to causal explanations.

The argument by which the defenders of Aristotle�s thesis discharge this

burden is roughly that one can account for an intentional action�s having a

point only by taking the action as fulfilling a certain function. That is, the

action has a point because the agent, being a rational agent, is conceived of

as a seeker or pursuer of what is good and not just an organism whose


behavior is due to the forces acting on it. Its point, then, is given in how the

action serves the function of his rational agency, which is to say, how it

contributes to his seeking or pursuing what is good. The forces acting on a

person at the time of action, the events and conditions immediately preced-

ing it, cannot by themselves account for this contribution. For accounting

for it presupposes the existence of a function that the action is fitted to

serve, and to explain the action as intentional by citing this function and

how the action is fitted to serve it is to explain the action teleologically and

not causally.

This defense of Aristotle�s thesis rests on the supposition that the point of

an intentional action is defined by a feature of the action thatmakes it seem

desirable or worthwhile to the agent. In the example of opening a window,

the feature that made it seem desirable to the agent was its making the

room more comfortable by removing a barrier to fresh air�s coming into it.

This feature defines the point of the action because it shows how the action

fulfills the function of rational agency, to seek or pursue what is good. The

question, then, that this defense raises is whether one can understand the

point of an intentional action without supposing that it is defined by a

feature of the action that makes it seem desirable or worthwhile to the

agent. If one can, then one need not suppose rational agency has this

function to explain intentional action. The defense would then fail. And if

the alternative way to understand an intentional action�s having a point

were consistent with causal explanations of action, then continued support

for Aristotle�s view would be nothing but exercises in anachronism.

6. The eliminability of teleological explanations

How, then, can we understand an intentional action�s having a point without

supposing that the function of rational agency is to seek or pursue what is

good? To answer, let us begin with a truism. The point of an intentional action

is to achieve its end.Whenone acts intentionally, one actswith foresight of the

events and states that are more or less likely to come about as a result of one�s

acting as one does, and one of these or some combination of them is what one

intends to bring about. This is the action�s end. Consequently, the feature of

the action that defines its point is its being more or less likely to result in this

event, state, or combination of events and states. The action, to be sure, may

therefore appear desirable or worthwhile to the agent in the sense that it gets


him what he wants. But it does not follow that it appears desirable or worth-

while to him in the sense that it helps him attain what is good, where what is

good is something independent of what he desires. Hence, if we can under-

stand how a person�s actions can come to have ends without supposing that

the person regards those ends as good, thenwe can understand the point of an

intentional action without supposing that its agent, as a rational agent, is a

seeker or pursuer of what is good.

What we need to determine is how a person�s actions originally acquire

their ends. Let us start with the earliest of human desires, those that develop

in infancy from the appetites and emotions that are already manifest in the

first few weeks after birth. These would be desires for satiation, sleep,

affection, relief of distress, among other states. Their satisfaction is a source

of pleasure, their frustration a source of discomfort or pain. These experi-

ences of pleasure and pain, when the desires are satisfied and frustrated,

come to be associated with the things whose consumption or use brings

satisfaction or whose opposition brings frustration. Accordingly, a child

takes pleasure in or is upset by having these things or by their presence. It

then acquires new desires and aversions, namely, desires for the things

whose consumption or use brings satisfaction and aversions to the things

whose opposition brings frustration. The new desires are for possession of

the objects for their own sake and not merely for their usefulness in getting

other things that one desires. Similarly, the new aversions are aversions to

the objects in themselves and not to the frustrations that those objects

cause. A child may come to be attached to a blanket or a stuffed animal,

for instance, in virtue of its being an aid to sleeping. Initially, the blanket or

stuffed animal may not have interested the child, but it later comes to be

something the child desires to possess because it comforts the child when

the latter is left alone to sleep. The pleasure of being so comforted is then

associated with the blanket or stuffed animal, and as a result, its possession

comes to be an end of the child�s actions not only at bedtime but at other

times as well. It becomes, that is, even at times when the child is not

disposed or expected to sleep, the object of the child�s searches, requests,

and demands. The child seeks to have it for its own sake.

Mill, in the chapter ofUtilitarianism inwhich he offers his famous proof of

the Principle of Utility, appeals to essentially the same process to explain

how people come to desire money, power, fame, and virtue for their own

sake. Focusing on the love of money, he writes:


There is nothing originally more desirable aboutmoney than about any heap

of glittering pebbles. Its worth is solely that of the things which it will buy;

the desires for other things than itself, which it is a means of gratifying. Yet

the love of money is not only one of the strongest moving forces of human

life, but money is, in many cases, desired in and for itself; the desire to

possess it is often stronger than the desire to use it, and goes on increasing

when all the desires which point to ends beyond it, to be compassed by it, are

falling off. It may, then, be said truly, thatmoney is desired not for the sake of

an end, but as part of the end.9

So too with other �great objects of human life� like power and fame. He

concludes:

Life would be a poor thing, very ill provided with sources of happiness, if

there were not this provision of nature by which things originally indiffer-

ent, but conducive to, or otherwise associated with, the satisfaction of our

primitive desires, become in themselves sources of pleasure more valuable

than the primitive pleasures both in permanency, in the space of human

existence that they are capable of covering, and even in intensity.10

Clearly, a process by which something comes to be desired for its own

sake as a result of repeated pleasurable experiences of it does not require

one�s regarding it as good in order for one�s possessing it to become an end

of one�s actions. It would be natural of course to regard it as good, but doing

so is inessential to this process. And similar things can be said of the process

of becoming averse to something as a result of repeated unpleasant experi-

ences of it. Further, it is worth noting that in Mill�s example of the love of

money, money originally serves as an instrument of obtaining other things.

As such, the lover of money will naturally see it as good, for being useful is a

form of being good. But to see money as useful is not to see it as an end.

Hence, his seeing it as good is independent of its being an end for him. That

is, he sees it as useful but also desires to acquire it for its own sake.What the

lover of money develops is an emotional attachment to money just as the

child in my example develops an emotional attachment to the blanket or

stuffed animal. The development marks the difference between desiring

something for its own sake and seeing it as useful to one�s attaining some-

thing else that one desires. Such attachments are the source of motives for

9 Mill, Utilitarianism, ch. 4, par. 6. 10 Ibid.


action, for they are, in effect, emotional dispositions. That is, they are

dispositions to have a range of different emotions with respect to the object

or person to whom one is attached. The different emotions in this range

then correspond to different motives: fear, anger, hope, jealousy, yearning,

and so forth. This account of how something becomes the end of one�s

actions is therefore consistent with causal explanations of those actions.

Finally, nothing about the process precludes something�s becoming the

end of a person�s actions, though he regards it as bad or evil. If a person

repeatedly gets pleasure from damaging or spoiling things, then he may

come to desire damaging or spoiling things for its own sake. Perhaps

Dostoyevsky�s underground man acquired his predilection for smashing

things in this way. Be this as it may, the point is that the ends of action

that result from this process may be such that one who desired them for

their own sake could also regard them as bad. We can thus see how the

underground man represents such an effective counterexample to the

Aristotelian model.

7. Modern skepticism about practical reason

Modern philosophers abandoned teleological explanations of natural phe-

nomena. They concentrated instead on causal explanations. Treating

human actions as natural phenomena, they looked for causal explanations

of them. That is, they sought to explain human actions as the effects of the

events and conditions that precede them. This way of explaining human

actions puts pressure on the traditional understanding of them as the work

of reason operating in the practical sphere. Indeed, some modern philoso-

phers, notably David Hume (1711–76), having replaced teleological explan-

ations of human actions with such causal ones, went on to deny that actions

were ever thework of reason. Because actions are not related to their causes

as conclusions of an argument are related to the premisses fromwhich they

follow, actions do not, strictly speaking, result from the operations of

reason. The causes of human action, according to Hume, are appetites and

passions, desires and emotions, alongwith the perceptions of theworld that

excite them, and the relation of an appetite or passion to the action it

causes – the relation of fear, say, to fleeing – is not that of inference. Fear

moves its subject to flee; it does not argue in favor of fleeing. While one can

reconstruct a teleological explanation so that the good at which the action


aims argues in favor of doing the action, one cannot similarly reconstruct a

causal explanation so that an action�s causes argue in favor of doing it.

Consequently, Hume denied that the processes of reason ever directly

yielded action.

Hume reinforced this view by characterizing reason as being limited in

its reach. On Hume�s characterization, reason works entirely in the service

of its possessor�s efforts to understand theworld around him. It is the power

by which one determines whether one�s ideas of the world are true. Hume,

in so characterizing reason, followed Hobbes. In Leviathan, Hobbes had

declared, �Reason is but reckoning (that is, the adding and subtracting) of

the consequences of general names agreed upon for the marking and

signifying of our thoughts.�11 In the same vein, Hume restricted the powers

of reason to demonstrative and probabilistic reason, or as we now com-

monly say, deductive and inductive reasoning. In either case, the conclu-

sions of reason – the conclusions one draws in reasoning – are thoughts one

can express in complete sentences. Actions, appetites, and passions are

therefore never the conclusions of reason. Such a limited conception of

reason marks Hume as breaking from the traditional understanding of

reason as the power by which men and women ruled their appetites and

passions. Indeed, Hume brashly asserted the opposite. �Reason,� he wrote

in a passage often quoted by his admirers, �is and ought only to be the slave

of the passions.�12 His skepticism about the possibility of practical reason

was unremitting.

Of course, in characterizing Hume as an unremitting skeptic about the

possibility of practical reason I do notmean that he denied the possibility of

reasoning about practical matters. This much should be clear from the

foregoing discussion of practical reason. It should be clear, in other words,

that the notion of practical reason I am using in my characterization of

Hume is a peculiarly philosophical notion and not the notion youwould use

in everyday conversation. Thus, if, unfamiliar with the peculiarly philo-

sophical notion of practical reason, you were asked what the difference

was between practical and speculative thought, you would likely say some-

thing like, �Practical thought concerns practical matters like how late the

local grocery store stays open or whether the dark spot on the garage floor

11 Hobbes, Leviathan, ch. 5, par. 2.12 David Hume, A Treatise of Human Nature, bk. II, pt. 3, sec. iii.


means your car is leaking oil; whereas speculative thought concerns matters

of intellectual curiosity like how human beings came to be bipedal or what

explains the appearance of rainbows after storms.� Ordinarily, that is, we

think of the difference between the two as a difference in their subject

matter. Practical thought is about practical matters. Speculative thought is

about matters of intellectual curiosity. There is no basis for skepticism,

obviously, on this way of identifying different areas of thought. It is a

rough-and-ready division, serviceable perhaps for some purposes. At the

same time, it is no more illuminating of the nature of thought than similar

divisions between secular and religious thinking or scientific and humanistic

thinking. The samegoes for the distinction betweenpractical and speculative

reason. On their ordinary notions, skepticism about either of them does not

arise. But there is also nothing illuminating in the distinction.

It is the peculiarly philosophical distinction that gives rise to skepticism.

This is a distinction between modes of reasoning and not of subject matter.

Being a distinction between modes of reasoning, it is a distinction within

human psychology. Moreover, it is sharp. There are no gray areas between

the two modes. Practical and speculative reason, on the philosophical dis-

tinction, are different inferential processes. The inferences that character-

ize speculative reason typically lead to beliefs as their conclusions. The

inferences that characterize practical reason typically lead to actions as

their conclusions. Take as an example of an exercise of speculative reason,

your inferring, from your observing wet streets and sidewalks in the morn-

ing, as you leave your home for work or school, that it rained during the

night. Your observations are judgments about the wet conditions of the

streets and sidewalks near your home, and from these judgments you

conclude that it rained the night before. The conclusion is a belief that

you form by inference from these judgments. Consider, in contrast, as an

example of practical reason, your taking an umbrella with you, as you leave

your home, because upon looking outdoors you see dark clouds overhead

and have a presentiment of rain. Your observations of dark clouds and the

imminence of rain result in your taking an umbrella, provided of course

that you wish to keep dry and understand that without an umbrella for

protection you won�t. Your taking an umbrella is the conclusion of this bit

ofmeans-to-ends thinking. It is themeans to keeping dry you take following

observations of your circumstances made under the influence of a wish to

keep dry. Decisions too, in the sense of a resolve to do something, when


made prior to their being carried out, are conclusions of practical thinking.

In either case, actions or decisions, the inferential process that results in

them is distinct from that of speculative reason.

Aristotle, in his account of the role of reason in action, introduced what

he called the practical syllogism to show how actions result from reasoning.

A practical syllogism consists of a major premiss, which is a general princi-

ple about what it would be good to have, a minor premiss, which is a

judgment of how one can realize that good in the circumstances one

faces, and a conclusion, which is an action. Thus, one might affirm that it

would be good to add calcium to one�s diet, judge that milk was high in

calcium, and conclude by pouring oneself a glass of milk. Such deliberation

leading to action exemplifies the exercise of practical reason, on Aristotle�s

account. Hume rejects this model. What yields this action, on his view, is

not reasoning from such premisses as (1) it would be good to have more

calcium in one�s diet and (2) drinkingmilk will providemore calcium.What

yields the action is the desire for more calcium in one�s diet and ultimately

the desire to live longer or the fear of developing osteoporosis in old age.

The transition from this desire or fear to action is not, on Hume�s view, a

transition of reason. It is not an inference. To deny the possibility of prac-

tical reason is to deny the possibility of such transitions of reason.

8. Hume�s meta-ethics

What meta-ethical views follow from Hume�s skepticism about practical

reason? To answer, let us replace, as the conclusion of the practical syllo-

gism we just considered, the action of pouring a glass of milk with the

directive �I should have a glass of milk� and then ask whether Hume�s

skepticism applies to the reasoning this new syllogism represents. Plainly,

the answer depends onhowone understands the nature of this directive. On

the one hand, if one takes it as an expression of an idea that could be true in

some circumstances, then it could be the conclusion of the reasoning the

new syllogism represents. In that case, the reasoning would be no different

from the deductive reasoning represented in an ordinary syllogism such as:

(3) New York is east of Philadelphia.

(4) I live in New York.

Hence, I live east of Philadelphia.


It would not, that is, be reasoning that issued in action. On the other hand,

if one holds that the directive necessarily guides action in the sense that if

one clearly comprehended and sincerely accepted it, one would necessa-

rily be moved to act accordingly, then one cannot by sound reasoning

infer it from (1) and (2). This is because the directive, on this second way of

understanding its nature, expresses a motivational state of mind, and

given skepticism about practical reason, the transition from the ideas

expressed by the premisses to such a state cannot be one of reason. In

short, one must have settled on one or the other understanding of the

nature of directives before any definitive meta-ethical view follows from

Hume�s skepticism.

Neither alternative, however, is wholly satisfactory. To conceive of direc-

tives as expressing ideas that could be true in some circumstances is to

conceive of them as comparable to factual statements. Yet as Hume himself

noted, nothing about factual statements as such moves a person to act.

Someone who was totally indifferent to the stated fact would be unmoved by

such a statement though he fully agreed with it. Only if the fact pleases or

displeases him, only, that is, if it raises his hopes or increases his fears, will his

apprehending it dispose him toward one or another course of action.

Directives, however, affect those who agree with them. They guide their

actions. Hence, a conception of them as comparable to factual statements

misrepresents their character.

Consider, then, the alternative way of understanding their nature, the

conception of them as necessarily action guiding. To conceive of them in

this way is to conceive of them as expressing motivational states. These are

states of appetite and passion, desire and emotion, and such states, no less

than actions, invite causal explanations according to which they are

brought about by antecedent events and conditions. Yet it would be odd,

at best, to explain a person�s coming to accept or reject a directive by citing

prior events and causes. Rather one often accepts or rejects directives in

view of considerations that argue for or against them. Someone concerned

about the brittleness of her bones, for instance, may be persuaded to drink

milk on learning that it supplies calcium that will strengthen her bones.

This shows that reasoning from premisses about matters that interest or

concern one can yield directives as conclusions. A conception of directives

as expressing motivational states therefore fails to capture how they can be

supported by reason and argument.


Hume himself opted for this second alternative. Because ethics is a

practical discipline, he argued, its teachings are meant to influence the

behavior of those who receive and accept them. Accordingly, he wrote,

�Morality . . . is more properly felt than judg�d of.�13 A directive to help

one�s neighbor or keep one�s promises expresses a sentiment of approval

toward that action; a directive not to lie or take more than one�s fair share

expresses a sentiment of disapproval. Arguments in support of such direc-

tives, Hume held, were arguments to establish facts whose apprehension

caused the sentiments of approval or disapproval the directives expressed.

That is, he held that men and women were so constituted that upon appre-

hending facts that pleased them, they felt approval toward actions that

brought about such facts and similarly for facts that displeased them. He

thus accounted for the place of reason and argument in ethical discourse. Or

so he thought.

The problem with his account, however, is, as we just noted, its reliance

on cause and effect to explain how we come to see certain facts as arguing

for or against directives. For the relation of cause to effect does not explain

the connection that holds between a fact and a directive when the former

argues for the latter. To see a fact as arguing for a directive means that you

cannot consistently affirm this fact and reject the directive unless you see

other facts that argue against the directive. If you know that milk strength-

ens your bones, for instance, and take this fact as arguing for the directive �I

should drink a glass of milk�, then it would be inconsistent for you to reject

this directive unless you saw other facts, such as your being allergic to dairy

products, as arguing against it. The reasonwhy youwould be inconsistent in

rejecting the directive lies in the basis on which you judge that the fact

about milk�s strengthening your bones argues for it. Recall that on Hume�s

view a fact can argue for or against a directive only if one is not indifferent to

it. In this case, the fact argues for the directive because you are not indif-

ferent to having stronger bones. To the contrary, you want to strengthen

your bones. Consequently, you cannot consistently desire to have stronger

bones and reject directives to take means to strengthening them unless

there are facts that argue against your taking those means. And in general,

one would be inconsistent in desiring x and seeing that to have x requires

doing y but rejecting the directive to do y, unless one saw facts that argued

13 Ibid., bk. III, pt. 1, sec. ii.


against one�s doing y. By contrast, if your sentiment of approval toward

drinking a glass of milk is merely the effect of your apprehending with

pleasure the fact that drinking milk will strengthen your bones, then there

would be no inconsistency in your feeling disapproval toward drinking a

glass of milk upon apprehending with pleasure that it would strengthen

your bones. It would be no more inconsistent than your falling asleep upon

drinking several cups of very strong coffee. Such an eventwould be contrary

to what drinking strong coffee causes, but there would be no inconsistency

in its occurring after you drank several cups of coffee. By the same token,

there would be no inconsistency in your rejecting the directive to drink a

glass of milk upon apprehending with pleasure the fact that drinking milk

will strengthen your bones. It is evident, then, that Hume�s account of the

place of reason and argument in ethical discourse fails.

9. Practical reason in modern philosophy

What does the failure of Hume�s account show about his skepticism con-

cerning practical reason? One obvious lesson is that Hume goes too far

when he denies that actions can be the conclusions of reason. One may, of

course, define reason, as Hume and Hobbes did, so as to exclude from its

operations any process of thought that issues in action. One may, that is,

restrict its processes to those of deductive and inductive reasoning or to the

�adding and subtracting� of words. But such restrictions are essentially

decisions to define �reason� narrowly. If one thinks, as Hobbes and Hume

apparently did, that reason is best conceived of as that power of the under-

standing whose workings are the subject of logic and mathematics, then

one will so restrict one�s use of the term. But if one thinks logic and

mathematics reflect only one side of reason, one will then define the term

more broadly. Accordingly one may, in light of considerations like those

that showed Hume�s account to be inadequate, decide to include within the

term�s scope deliberation issuing in action. Those considerations showed

the possibility of inconsistency between one�s aims and one�s actions and

therefore, if one thinks reason is best conceived of as including the cogni-

tive power by which one discerns and corrects inconsistency in one�s

thought and action, onewill extend the scope of the term �reason� to include

both spheres. One will, in other words, add practical reason to the powers

that fall under the general concept of reason.


Consequently, to defend Hume�s skepticism concerning practical reason

one must moderate its categorical denial of the existence of inferences of

reason whose conclusions are actions. If one wishes, that is, to continue to

deny the existence of such inferences, one must then acknowledge the

narrow sense in which one is using the term �reason� and, hence, the

possibility of a broader sense of the term that includes practical reason

within its scope. At the same time, one�s acknowledgement of a broader

sense does not mean that one is abandoning Hume�s skepticism in its

entirety. For although Hume�s argument against the possibility of there

being inferences of reason whose conclusions are actions fails to establish

this view categorically, there is implicit in the argument a skeptical argu-

ment against a different thesis about practical reason, and this implicit,

skeptical argument has proven hard to answer.

The thesis it opposes takes us back to the dispute between rationalists

and naturalists thatwe considered in chapter 3. Recall that the question that

divides these two schools is whether reason alone can initiate action.

Rationalists believe that it can, and accordingly attribute to reason the

power to produce motives that conflict with the motives arising from

animal appetite and passion. Naturalists hold, to the contrary, that all

motives arise from animal appetite and passion and therefore reason�s

role in action is limited to informing its possessor about the consequences

of the actions he is moved to do so that he acts on enlightened motives.

Hume in arguing against the possibility of practical reason makes a strong

case against the rationalists� hallmark thesis that reason alone can initiate

action and for the naturalists� contrary view. This case, Hume�s anti-

rationalism, is what remains of his skepticism about practical reason

once, having allowed for a broad sense of �reason,� one accepts that actions

can be the conclusions of inferences of reason in this sense. Let us call this

the modified Humean view.

Hume�s case consists in his observing that none of the different opera-

tions of reason generates motives of action. Hume, as we noted, divides the

operations of reason into those of demonstrative and probabilistic reason-

ing. With respect to the former, he observes that because it concerns

abstract relations of ideas, it never goes further in its conclusions than

ideas. When we reason deductively, we begin with premisses and pass by

inferences to a conclusion either by virtue of the necessary connection

between propositions that have the form the premisses have and a


proposition that has the form the conclusion has, or by virtue of the

necessary connection between concepts contained in the premisses and

concepts contained in the conclusion. In either case the reasoning as such

stays within our understanding or system of beliefs. With respect to the

latter, Hume concedes that because it may concern relations of cause and

effect, if we have in mind the prospect of something pleasurable or painful,

our thoughtswill pass, guided by reasoning from effect to cause, to events of

a type that would cause it, and we may then be moved by desire to bring

about such events or by aversion to avert them. But, Hume then argues, in

this case the motive of desire or aversion is due to the pleasure or pain in

prospect and not the reasoning that leads us to think of its cause. Indeed, if

we neither took pleasure in the prospect of gaining this object nor found it

painful, we would not be moved either by desire or aversion in contemplat-

ing its cause. �Where the objects themselves,� Hume writes, �do not affect

us their connexion can never give them any influence; and �tis plain that as

reason is nothing but the discovery of this connexion, it cannot be by its

means that the objects are able to affect us.�14

Suppose, then, we add practical reason toHume�s list of the operations of

reason. Clearly, this addition will not affect Hume�s anti-rationalism if we

conceive of practical reason as the power to discern and resolve inconsis-

tencies between one�s aims and one�s tendencies to act contrary to those

aims. It should be clear, that is, that Hume could say the very same thing

about practical reason, conceived of in this way, as he does about demon-

strative and probabilistic reason. For the power to discern and resolve

inconsistencies between one�s aims and one�s tendencies to act contrary

to those aims operates on already existing motives, and there is no need to

suppose that it generates any newmotives to understand its operations. You

may be inclined, for instance, to take the scenic route on a trip to visit your

parents, but if you intend to arrive at your parents� home in time for dinner

and you realize that taking the scenic route requires more time than you

have, then your interest in having dinner with your parents will prevail. No

additional motive is necessary to resolve the conflict. Or you may be plan-

ning a vacation and deciding between the Canadian Rockies, where you

could ski, and Baja California, where you could surf. You can�t do both and

therefore youmust choose according to which fits best with your larger aim

14 Ibid., bk. II, pt. 3, sec. iii.


of having a pleasurable vacation. Again, no additional motive among the

desires to ski, surf, and have a pleasurable vacation is necessary to resolve

the conflict.

The modified Humean view, while weaker than Hume�s full-blown skep-

ticism about practical reason, is nonetheless sufficient to enable him to

maintain the meta-ethical view about the roots of morality that we touched

on in the previous section, the view that morality is at bottom a matter of

feeling and not reason. His anti-rationalism is one of two propositions on

which he bases this view. The other is that ethics is an essentially practical

discipline. Its teachings are nomere statements of fact but are rather guides

to conduct that should move all who understand and accept them to follow

them. Ethics is the study of directives, and directives guide action. Since

reason is inert, as Hume liked to say in expressing his anti-rationalism, the

teachings of ethics must speak to our desires and emotions if they are to

move us. Hence, it follows, on Hume�s view, that there is no directive,

specifically no moral precept or dictate about what a person ought to do,

the rejection of which is necessarily contrary to reason. Whether rejection

of a moral precept goes against reason depends on the interests, desires,

cares, fears, and the like of the personwho rejects the precept, for only if the

rejection is inconsistent with the pursuit of some end he is disposed to

pursue will it be contrary to reason. Hume�s meta-ethical views therefore

allow dissent from moral dictates about right and wrong or what ought or

ought not to be done that one cannot criticize for being irrational or even

untrue. On this point, Hume�s views are compatible with those of the

defenders of existentialist ethics and opposed to those of Aristotelians.

The opposition between the Aristotelian view and the modified Humean

view on the question of the truth of moral precepts and dictates nicely

brings out the difference in conceptions of practical reason. To illustrate,

consider a simple directive like �You ought to drive more slowly.� On the

Aristotelian view of practical reason, this directive, like the very action of

driving more slowly, follows from your affirming that safety from auto

accidents is a good and seeing that slowing the speed of your car will

make you safer in the current conditions of the road on which you are

driving. Safety from auto accidents is plainly a good in almost any circum-

stance of driving you can imagine, and hence the directive is true if driving

more slowly will make you safer. Of course, if drivingmore slowly would be

hazardous because, for instance, you are driving in heavy, fast-moving


traffic and already driving slower than the other cars around you, then the

directive is false. By contrast, on themodifiedHumean view, the truth of the

directive depends, not on there being a good that one could realize by

driving slowly, but rather on one�s having an end – that is, the object of

some desire or interest – that one would realize by so driving. For the

directive would not guide one�s actions unless following it advanced some

end one has. A driver who loves to speed and cares nothing for his own

safety or that of other drivers would reject this directive (provided of course

he had no other end that driving more slowly would help him realize), and

in doing so he would not be displaying any defect of reason or ignorance of

ethical truth. On the Aristotelian view, ethics is an objective study, and its

practical truths are determined through investigation of the human good.

On themodified Humean view, by contrast, ethics is a practical studywhose

results are not certain to be universal truths but rather precepts and dictates

whose truth is relative to the particular desires, emotions, and sentiments

of each individual.

On the modified Humean view, there are universal ethical truths only if

all men and women have in common certain desires whose objects com-

pliancewithmoral precepts and dictates would further. Needless to say, this

condition is not likely to obtain, for there are almost certainly a few adult

human beings whose personalities are sufficiently anti-social to make them

immune to the guiding force of these precepts and dictates. Nonetheless,

human beings may generally share a social nature relative to which moral

precepts and dictates are true, and it is upon the supposition of shared

human sociability that Hume makes out his argument for morality�s being

rooted in social feelings rather than reason. The precepts and dictates of

morality, on this view, are not universal truths. But they are generally true.

That is, for human beings generally they guide their actions.

10. Kant�s notion of practical reason

Kant�s practical philosophy stands in direct opposition to Hume�s on the

question of the roots of morality. Kant believes morality has its roots in

reason. His account of the Categorical Imperative as the supreme principle

of morality is testament to this belief. At the same time, like Hume, Kant

holds that moral precepts guide action in the sense that a personwho under-

stands and accepts them is moved to act accordingly. Hence, Kant takes


reason as having the power to initiate action aswell as to inform its possessor

about how to achieve his or her ends. Kant thus affirms the hallmark thesis of

rationalist ethics and, indeed, he advances his account of the Categorical

Imperative as the supreme principle of morality to vindicate it. Just as Hume

is the foremost defender of anti-rationalism in ethics among modern philos-

ophers, Kant is the foremost defender of rationalism in ethics.

Kant clearly sees that to vindicate the rationalist thesis requires an

expansion of our understanding of practical reason. If the operations of

practical reason are limited to our discerning and resolving inconsistencies

between our aims and our tendencies to act contrary to those aims, then its

role in the production of action will be that of an arbiter among conflicting

nonrationalmotives. In that case,moralitywould have nomore claim to our

allegiance than any other object of interest and desire. Consequently, the

very operations of reason that validate moral precepts and dictates must

also supply the motives to act as they direct. They must supply purely

rational motives. This is what Kant calls pure practical reason. He did not

think, however, that an exposition of morality alone showed that reason

was the source of thesemotives. His exposition consists of an analysis of the

idea of acting from duty or respect for moral law that yields the different

formulations of the Categorical Imperative as the supreme principle of

morality, but the analysis of an idea, as he observed, is not sufficient to

establish its realization in human lives or even how it could be so realized.

To establish either, one must find in our powers of reason operations

sufficient to bring about its realization.

How then can practical reason be reconceived to encompass the opera-

tions required for moral motives to be purely rational motives? These must

be operations that regulate practical thought just as the operations of theo-

retical reason regulate speculative thought. Rational action results from

practical thinking when it is properly regulated by practical reason.

Rational belief results from speculative thinking when it is properly regu-

lated by theoretical reason. This distinction is fundamental to Kant�s

philosophy. His program of replacing, as the foundations of moral law,

substantive standards of right and wrong that we know by an exercise of

rational intuitionwith formal principles of practical reason that regulate the

thinking we undertake in determining our actions could not proceed with-

out it. These formal principles must therefore include some – at least one –

whose regulation of practical thought yields motives without the help of


antecedently existing motives. Kant identifies two mutually exclusive types

of formal principle, hypothetical imperatives and categorical imperatives.

The former regulate practical thought by transmitting the motive to achieve

an end that the agent has resolved to pursue to those acts that the agent sees

are the necessary means to achieving it. They cannot, therefore, be the

source of purely rational motives. The latter regulate practical thought by

testing plans of action, what Kant calls maxims, against criteria of lawful-

ness and suppressing the inclination to act on those maxims that fail the

test. Such regulation may yield purely rational motives provided the testing

does not proceed under the influence of antecedently existing motives.

Kant maintains that it does not. When you test a maxim of action against

the criteria of lawfulness contained in the Categorical Imperative and

determine that the maxim fails, you are moved to suppress whatever incli-

nation to act on that maxim you have. Your motive to suppress the inclina-

tion is what Kant calls the motive of duty or respect for moral law. While

this motive opposes your inclination to act on the maxim, the conflict

between the two is not like the conflict in our earlier example between

your desire to take the scenic route to your parents� home and your desire to

arrive at their home in time for dinner. In this earlier example, the two

conflicting desires are on a par in the sense that you can resolve the conflict

by finding some common measure by which to decide between them. Kant

thought the common measure in all cases of such conflict was the contri-

bution to your happiness that satisfaction of the desire wouldmake. That is,

on Kant�s view, you decide between them by determining which desire is

such that its satisfaction would give you the most pleasure. But you needn�t

follow Kant on this point. You might use instead, as the common measure,

the importance you place on the desire�s being satisfied. Accordingly, you

would decide between them by determining which desire is such that its

satisfaction matters more to you. Be this as it may, Kant�s main point is that

the conflict in motives that occurs when one is moved by themotive of duty

to suppress an inclination to act unlawfully is different. One cannot resolve

it by finding a common measure by which to decide between them. To the

contrary, Kant holds, there is no common measure.15 The motive of duty,

15 Immanuel Kant, Critique of Practical Reason, L.W. Beck, trans. (Indianapolis, Ind.:

Bobbs-Merrill, 1956), pp. 29–30. Reference is to page numbers in the Preussische

Akademie edition.


being a motive to suppress an inclination to act, is of a higher order than

that inclination. Indeed, because it inheres in the judgment that interdicts

that inclination, the judgment that the maxim on which one is inclined to

act fails the tests of lawfulness contained in the Categorical Imperative, its

superiority to that inclination is manifest. Its manifest superiority, Kant

argues, reflects its origins in the operations of reason.

The argument is succinct. The suppression of inclination that results

from one�s recognizing that to act on the maxim to which the inclination

gives rise would be unlawful exemplifies a person�s capacity to step back

from the natural desires and emotions that move him and to assess their

influence. When the assessment moves one to suppress the inclination, its

motive force cannot originate in a natural desire or emotion. Rather it must

come from the exercise of reason in which the assessment consists. It

cannot originate in a natural desire or emotion because, if it did, one

could use some common measure to resolve the conflict. But the manifest

superiority of the motive of duty shows that there is no such measure. The

motive must therefore originate in reason. Kant thus locates the source of

purely rational motives in our capacity for reflective assessment of our

natural desires and emotions, when that assessment is not undertaken in

the service of some natural desire or emotion. This capacity is exemplified,

in Kant�s ethics, by the use of the Categorical Imperative to test the maxims

of one�s actions, but his argument for its being the site of pure practical

reason does not depend on the protocols of the test. What is crucial in the

argument is that people are capable of reflectively assessing the natural

desires and emotions on which they act and of being moved by those

assessments.

Of course, even on themodified Humean view, people can assess from an

unbiased standpoint the natural desires and emotions on which they act.

Indeed, Hume makes such assessments central to his account of virtue and

vice, for his account proceeds from the idea that motives are the proper

objects of the sentiments of moral approbation and moral disapprobation.

Furthermore, one experiences these sentiments from a general or disinter-

ested view of theworld, a viewpoint everyone ormost everyone is capable of

taking. They therefore have the character of reflective assessment. Yet the

assessment in each case consists solely in one�s being pleased or displeased

upon one�s discerning or thinking about the inner spring of a person�s

actions. As such it carries no distinctive authority relative to other natural


desires and emotions. These assessments are states of approval and disap-

proval, to be sure, but for Hume �approbation� and �disapprobation� are

merely names for certain pleasant feelings and certain unpleasant ones.

Consequently, when there is a conflict between the sentiment of disappro-

bation, say, and the inner spring that is its object, the conflict is similar to

other conflicts between natural desires. One can resolve it by applying a

common measure and deciding between them. It raises no question about

the superiority of disapprobation to the inner spring. Unlike Kant�s account,

there is therefore no reason to suppose that the assessments that the senti-

ments of approbation and disapprobation imply are products of reason.

What secures the rationalist thesis, then, on Kant�s account, is not merely

one�s capacity to assess reflectively from a disinterested standpoint one�s

motives, but moreover the nature of the standpoint from which one makes

these assessments. It must be one of disengagement from one�s natural

desires and emotions.

11. Freedom and reason

Kant identifies this disengaged standpoint with human freedom. Were we

unable to assess our natural desires and emotions from such a standpoint

and check those we found unacceptable, we would be at their mercy, so to

speak – and the mercy of the forces of nature generally – in all that we did.

That is, if our natural desires and emotions were the ultimate springs of all

human action, then every human action would be the result of those events

and conditions that produced the motive, and those events and conditions

would in turn be the product of prior events and conditions, and so on.

Human freedom, in Kant�s view, requires that men and women be capable

of acting on their own volition, and their power to determine their will

must be independent of the forces of nature. That power, Kant holds, is

found in pure practical reason. It is the power to conform one�s actions to

standards bywhich one�s natural desires and emotions, through assessment

of themaxims of action to which they give rise, are themselves assessed and

which issue from reason. These standards, then, are moral precepts whose

authority is universal and whose truth is absolute.

ThemodifiedHumean view, by contrast, does not attribute such freedom to

human beings. Naturalism, which this view represents, holds that human

actions are events of nature and as such are explainable by the forces and


conditions studied in the natural sciences and that explain other events in

nature. To the naturalist, the idea of a standpoint from which a person can

disengage from his or her natural desires and emotions is anathema. It is the

idea of a place, free from the forces of nature, to which men and women can

remove themselves, and there is no such place. The freedom it promises is

incoherent if human beings wholly belong to the natural world. Human free-

dom is, according to naturalism, the freedompeople enjoywhen they dowhat

they desire to do, provided that the desires they act on are desires by which

they are pleased to be moved. Freedom, on this view, is something to be

achieved in life, for it not only requires an external environment conducive

to one�s being able to dowhat onewants to do but also requires freeing oneself

of internal conflicts that create ambivalence and uncertainty about how to

conduct one�s life. And internal conflicts are often themost disabling factors in

one�s life as well as the hardest to overcome. While achieving internal coher-

ence requires practical reason to see the inconsistencies, their resolution does

not always come about solely through rational judgment about one�s ends and

goals. On this view, whether there are any directives that count as universal

practical truths depends onwhether there is any end or goal that every human

being has, and whether there is such an end or goal cannot be determined a

priori. In any case, all practical truths are relative to people�s ends. There are no

directives (and so no moral precepts) whose truth is absolute.

Existentialists, too, deny that any universalmoral precepts are absolutely

true. The doctrine of radical choice entails as much. Consequently, contrary

to Kant, they treat human freedom as incompatible with there being uni-

versalmoral directives that are absolutely true. At the same time, they agree

with Kant in seeing the capacity for freedom as essential to our humanity.

Freedom is not a condition human beings achieve. It is rather something

inherent in our power of choice. But in opposition to Kant they do not

identify that power with practical reason. It is a power one may exercise

in reflecting on desires and emotions that are the motives of one�s actions,

but one exercises it in endorsing or rejecting those motives according to

one�s decision as towhat sort of person to be andwhat kind of life to live and

not in assessing them against standards issued by reason. Human freedom,

on the existentialist view, is not coincident with reason. When one exer-

cises it in making the most fundamental choices as to how one should live,

one cannot rely on reason to determine for oneself what choices are correct.

Even themanifest superiority of themotive of duty to which Kant appeals is


open to question, on the existentialist view. Deferring to it as showing the

authoritative nature of moral standards for assessing one�s natural desires

and emotions would be abandoning one�s freedom, for the existentialist

view is that the ultimate authority in matters of how one ought to live and

thus what desires and emotions to act on is the agent�s own lucid and

sincere decisions. Thus, while the motive of duty may appear superior to

every natural desire or emotionwithwhich it conflicts, it is still subordinate

to one�s decision about how one ought to live. And while the followers of

Kant may then object that to go against the motive of duty is to go against

reason, the existentialists can respond that ignoring duty is still a coherent,

albeit radical choice, even if contrary to reason on Kant�s strengthened

conception of reason. It is therefore an intelligible exercise of personal

autonomy. To think that you may not ignore duty when doing so would

be coherent is, on the existentialists� view, to abandon your freedom.

Our study of practical reason in modern philosophy may appear, then, to

leave it to us to decide among these three opposing positions. And if so, itmay

then seem as though we would have to favor existentialism by default, for

deciding among these positions appears to require a choice forwhich there is

no standard of correctness. It appears, that is, to require a radical choice.

How, after all, can one decide which of the three conceptions of freedom that

correspond to these three positions correctly applies to humankind? Such a

decision appears nothing short of deciding how to understand one�s own

humanity and the humanity of others, and what could be a more fundamen-

tal and yet indeterminate decision than that? Surely a study that ends by

leaving things open to such a decision implies the primacy of existentialism.

Yet such a conclusion would be too hasty. We must not mistake a seem-

ingly indeterminate philosophical question for a first-order question of

ethics, and the questions giving rise to radical choice are of the latter

kind. What our study discloses, rather, is that the question of the nature

of human freedom is itself a second-order question, a question about

whether we can find answers based in reason to first-order questions of

ethics and what sort of answers they are if we can. It is a question of meta-

ethics. To be sure, it may be one for which we will never have enough

knowledge about and understanding of our humanity to arrive at a deter-

minate answer. But that possibility should not foreclose our continuing to

debate it. Like other major problems in philosophy, only further reflection

and argument will tell whether there is a bottom to its depths.


Appendix: Diagram of different teleologicaltheories

*Perfectionistic utilitarianism, like classical utilitarianism, admits of act and rule versions.

Outlook ofself-concern

Egoism(happiness)

Eudaimonism(well-being)

Perfectionisticegoism

Hedonisticegoism =hedonisticeudaimonism

Perfectionisticeudaimonism

Platonic eudaimonism

Plato’s rationalism

Outlook of impartiality

Aristotle’s naturalism

Contemporary eudaimonism

Perfectionisticutilitarianism*

Act utilitarianism

Ruleutilitarianism

Classical (hedonistic)utilitarianism

Teleological theories

233

Works cited

Aquinas, Thomas. Summa Theologica.

Aristotle. Nicomachean Ethics.

Bentham, Jeremy. An Introduction to the Principles of Morals and Legislation.

Butler, Joseph. Fifteen Sermons Preached at the Rolls Chapel.

Camus, Albert. The Myth of Sisyphus and Other Essays, J. O�Brien, trans. (New York:

Vintage Books, 1955).

Carroll, Lewis. Through the Looking Glass (and What Alice Saw There).

Dostoyevsky, Fyodor. Notes from Underground/The Double, J. Coulson, trans.

(Harmondsworth, England: Penguin Books, 1972).

Epicurus. �Principle Doctrines.�

Forster, E.M. Two Cheers for Democracy (New York: Harcourt Brace, 1951).

Grotius, Hugo. On the Law of War and Peace, F.W. Kelsey, trans. (Oxford: Oxford

University Press, 1925).

Hobbes, Thomas. Leviathan.

Hume, David. A Treatise of Human Nature.

James, William. The Principles of Psychology, 2 vols. (1890; reprinted New York:

Dover Publications, 1950).

Kant, Immanuel. Critique of Practical Reason, L.W. Beck, trans. (Indianapolis, Ind.:

Bobbs-Merrill, 1956).

Groundwork of the Metaphysics of Morals, H. J. Paton, trans. (New York: Harper &

Row, 1964).

Mill, John Stuart. Utilitarianism.

Moore, G. E. Principia Ethica (Cambridge: Cambridge University Press, 1903).

Plato. Republic.

Prichard, H. A. �Does Moral Philosophy Rest on a Mistake?� Mind 21 (1912):

21–37.

Sartre, Jean-Paul. �L�existentialisme est un humanisme�; reprinted as

�Existentialism Is a Humanism,� Philip Mairet, trans., in Existentialism from

Dostoevsky to Sartre, Walter Kaufmann, ed. (Cleveland: TheWorld Publishing

Co., 1956).

234

Sheinwold, Alfred. 5 Weeks to Winning Bridge, rev. edn. (New York: Pocket Books,

1964).

Sidgwick, Henry. The Methods of Ethics, 7th edn. (London: Macmillan and Co.,

1907).

Twain, Mark. The Adventures of Huckleberry Finn.

Works cited 235

Suggested further readings

Chapter 1

Bennett, Jonathan. �The Conscience of Huckleberry Finn,� Philosophy 49 (1974):

123–34.

Darwall, Stephen. The Second Person Standpoint (Cambridge, Mass.: Harvard


Ross, W.D. The Right and the Good (Oxford: Clarendon Press, 1930).

Strawson, P. F. �Social Morality and Individual Ideal,� Philosophy 36 (1961): 1–17.

Chapter 2

Broad, C.D. �Egoism as a Theory of Human Motives,� Hibbert Journal 68 (1950):

105–14; reprinted in Broad�s Critical Essays in Moral Philosophy, David Cheney,

ed. (London: George Allen & Unwin Ltd, 1971), pp. 247–61.

Gauther, David. �Assure and Threaten,� Ethics 104 (1994): 690–721.

Kavka, Gregory. Hobbesian Moral and Political Theory (Princeton, N.J.: Princeton


Sidgwick, Henry. �Egoism,� in The Methods of Ethics, 7th edn. (London:

Macmillan & Co., 1907), bk. II, pp. 119–95.

�Pleasure and Desire,� in The Methods of Ethics, bk. I, ch. 4, pp. 39–56.

Chapter 3

Annas, Julia. Introduction to Plato�s Republic (Oxford: Clarendon Press, 1981).

Griffin, James. Well-Being: Its Meaning, Measurement and Moral Importance (Oxford:

Clarendon Press, 1986).

Kraut, Richard. Aristotle on the Human Good (Princeton, N.J.: Princeton University

Press, 1989).

Plato. Gorgias.

Sachs, David. �A Fallacy in Plato�s Republic,� Philosophical Review 72 (1963): 141–58.

236

Chapter 4

Kagan, Shelly. The Limits of Morality (Oxford: Clarendon Press, 1989).

Lyons, David. Forms and Limits of Utilitarianism (Oxford: Clarendon Press, 1965).

Parfit, Derek. Reasons and Persons (Oxford: Clarendon Press, 1984).

Railton, Peter. �Alienation, Consequentialism, and the Demands of Morality,�

Philosophy & Public Affairs 13 (1984): 134–71.

Rawls, John. �Two Concepts of Rules,� Philosophical Review 64 (1955): 3–32.

Urmson, J.O. �The Interpretation of theMoral Philosophyof J. S.Mill,� Philosophical

Quarterly 3 (1953): 33–39.

Chapter 5

Adams, Robert M. �A Modified Divine Command Theory of Ethical Wrongness,�

in Religion and Morality: A Collection of Essays, Gene Outka and John P. Reeder,

Jr., eds. (Garden City, N.J.: Anchor Press, 1973), pp. 318–47.

Herman, Barbara. �Mutual Aid and Respect for Persons,� Ethics 94 (1984): 577–602.

O�Neil, Onora. Acting on Principle: An Essay on Kantian Ethics (New York: Columbia


Ross, W.D. The Right and the Good (Oxford: Clarendon Press, 1930).

Schneewind, J. B. The Invention of Autonomy (Cambridge: Cambridge University

Press, 1998).

Strawson, P. F. �Ethical Intuitionism,� Philosophy 24 (1949): 23–33.

Chapter 6

Camus, Albert. The Stranger, Stuart Gilbert, trans. (NewYork: Alfred Knopf, 1946).

Hill, Thomas E., Jr. Dignity and Practical Reason in Kant�s Moral Theory (Ithaca,

N.Y.: Cornell University Press, 1992).

Korsgaard, Christine. �The Right to Lie: Kant on Dealing with Evil,� Philosophy &

Public Affairs 15 (1986): 325–49.

Nagel, Thomas. �The Absurd,� Journal of Philosophy 68 (1971): 716–27.

Sartre, Jean-Paul. Being and Nothingness, Hazel Barnes, trans. (NewYork:Washington

Square Press, 1956).

Wood, Allen.Kant�s Ethical Thought (Cambridge: Cambridge University Press, 1999).

Chapter 7

Anscombe, G. E.M. Intention, 2nd edn. (Ithaca, N.Y.: Cornell University Press,

1963).

Suggested further readings 237

`rdal, PÆll S. Passion and Value inHume�s Treatise, 2nd edn. (Edinburgh: Edinburgh


Davidson, Donald. �Action, Reasons, and Causes,� Journal of Philosophy 60 (1963):

685–700.

Frankfurt, Harry. �FreeWill and the Concept of a Person,� Journal of Philosophy 68

(1971): 5–20.

Korsgaard, Christine. �Skepticism about Practical Reason,� Journal of Philosophy

83 (1986): 5–25.

Nussbaum, Martha. The Fragility of Goodness: Luck and Ethics in Greek Tragedy and

Philosophy (Cambridge: Cambridge University Press, 1986), ch. 9.

Stocker, Michael. �Desiring the Bad,� Journal of Philosophy 76 (1979): 738–53.

Velleman, J. David. �The Guise of the Good,� Nous 26 (1992): 3–26.

Wiggins, David. �Truth, Invention, and the Meaning of Life,� Proceedings of the

British Academy 62 (1976): 331–78.

Williams, Bernard. �Internal and External Reasons,� in Rational Action,

Ross Harrison, ed. (Cambridge: Cambridge University Press, 1980),

pp. 17–28; reprinted in B. Williams, Moral Luck (Cambridge: Cambridge

University Press, 1981), pp. 101–13.

238 Suggested further readings

Index

action

intentional, 33�36, 38, 39

motiveless, 205, 206�7

appetite, 63�64, 65, 71, 77, 79, 80, 142, 211,

214, 216, 217

Aquinas, Thomas, 205

Aristippus of Cyrene, 60n

Aristotle, 81�86, 90�91, 123, 194, 201, 202,

203�05, 219, 225�26

autonomy

of the will, 168, 169

personal, 178, 180, 183, 185�87, 188, 192,

194, 195, 197, 203�4, 232

aversion, see Desire

bad faith, 187, 188, 191

Bentham, Jeremy, 94�6, 103, 107, 111, 113

Butler, Joseph, 40�42, 70, 210

Camus, Albert, 192�94, 198, 204�05

Carroll, Lewis, 193

categorical imperative, 144�56, 158, 161,

175, 176, 189, 226�29

CI procedure, 147�56, 160, 163, 165, 166�7,

168, 171�72, 174�176, 178, 179, 180, 189

equivalence of different formulations,

161�62

formula of autonomy, 168�69, 172�73

formula of humanity, 161�67

formula of universal law, 146, 161�62,

163, 166

choice

doctrine of radical, 184�86, 190�92, 193,

194, 196�97, 199�201, 203, 206, 231

Christian ethics, 15, 97, 123�24

Christianity, 123�24, 129, 131, 140

Cicero, 123

conscience, 123�24

consequentialism, 101�03, 115�16

cooperation, 45�48, 51, 54

courage, 72�73, 80, 86

cruelty to animals, 52�53

death, 1, 78

Deep Blue, 117�18

Descartes, Rene, 192

desire, 39�42, 79, 142, 216, 224

dignity, 177�78

directives, 198�201, 219�22, 225�26

divine command theory, 128�30

Dostoyevsky, Fyodor, 202�3, 207�9, 216

duality of human nature, 159, 166,

178, 179

duty (obligation), 12�14, 20, 231, 232

egoism, 25�27, 30, 32�37, 39, 42�44, 46, 50,

51, 52, 53, 55, 56�58, 59, 93, 125

end-in-itself, 160�62

Epicurus, 25, 61, 64�65, 71, 72, 123

equality

in Hobbes, 45�46

under law, 127�28

ethics

analogy to mathematics, 136�40

defined, 7�8

deontological conception of, 14�19,

21�22, 93, 112, 124�27, 170�71, 195

ground-level questions of, 197�99

239

ethics (cont.)

teleological conception of, 14�15, 25,

123, 124, 195, 201, 233

theories of, 14�16, 22�24

Euclid, 137�38, 139, 162

eudaimonism, 56�60, 64, 71�77, 81�92, 93,

201, 202, 203�5

contemporary, 90�92

Platonic formof,58�60,71�77,90�91,202n

evaluative statements, see valuations

existentialism, 181�88, 189�95, 196�97,

200�01, 203, 206, 231�32

explanation

causal, 211, 212�13, 216�17, 220

teleological, 211�13, 216

formalism, 145, 147�56, 157, 163, 171�72

Forster, E.M., 184

free-rider

problem of, 46�48, 53, 55

function argument, 84�90

Godwin, William, 98

Greatest Happiness Principle, see Principle

of Utility

Grotius, Hugo, 132�35

happiness, 1, 28�33, 36, 38�43, 51, 54,

152, 159

hedonic calculus, 95�97, 114

hedonism, 29�31, 55, 56, 57�58, 59, 60�63,

83, 95

highest good, 25, 32, 34, 47, 55, 57�58, 61,

62, 71, 93, 124, 126, 127, 132

Hobbes, Thomas, 25, 33, 44�53, 54�55, 125,

128�29, 150, 155, 199�201, 217, 222

honesty, 1�7, 11�12, 14, 16�19, 22�23,

50, 86, 87, see also justice

Hume, David, 216�17, 219�27, 229�30

hypothetical imperative, 143�45, 175, 176,

179, 200n, 201, 228

ideals, 23�24

moral, 23�24

of self-government, 168, 179

impartiality, 96�99, 107

imperative, 142�46, see also categorical

imperative, hypothetical imperative

instrumental good, 32

integrity, 186�88, 192�93

intellectualism, 132�35

James, William, 30

justice, 4�7, 10�11, 13�14, 21�22, 58�59,

73�75, 76, 80, 86, 87, 90, 92, 99�108,

110�11, 117�20, 132, 134, 135, 196,

202, 208

Kant, Immanuel, 140�56, 157�81, 186, 189,

199, 226�32

Kasparov, Gary, 117�18

kingdom of ends, 169�71, 172�73, 180

Leibniz, Gottfried, 132

meta-ethics, 197�201, 219�22, 225, 232

Mill, J. S., 60, 65�70, 97, 103�12, 113, 117,

118, 214�15

moderation, 73, 80, 86

Moore, G. E., 67

moral community, 19�22

moral law, 124�35, 140, 141�44, 157�58,

160, 169, 175, 186

moral order, 125�26, 127, 129, 140

moral psychology

theories of, 77, 78

moral skepticism, 53�54, 55

morality

as a universal ideal, 10

authority of, 10, 196

bindingness of, 102�03, 108

conventional, 8�10, 52

deontological conception of, 14, 123�28

public character of, 120�21

roots of, 221, 225, 226

teleological conception of, 14

motives, 33�43, 45, 77�80, 146, 170�71,

179, 182�83, 201, 205, 208, 209�11,

215�16, 223�25, 227

and right action, 115�16

of duty, 168�69

of injustice, 75�76, 90-91

240 Index

natural law, 44�45, 47, 52, 170

scholastic conception of, 124, 126

naturalism, 77�86, 211, 223, 230�31

nature

teleological conception of, 212

normative statements, see directives

ought, 12�14, 15, 17

Paul, 123�24

perfection, 72, 76�72

perfectionism, 32, 56�57, 59, 60, 66, 69, 70,

81, 83

personality, 71

well-ordered, 63, 71�77, 79, 80

Plato, 4�7, 55, 58, 59, 60�61, 62�65, 71�77,

79, 90, 91, 123, 135

pleasure, 1, 37�41, 55

high-grade v. low-grade, 66�70

positive law, 125�26

practical reason, 140�46, 194, 195, 211,

216�32

skepticism about, 219�24

v. speculative reason, 217�19

practical syllogism, 219

Principle of Utility, 94�108, 110, 133, 188,

189, 214

psychological egoism, 33�37, 42, 199�201

psychological hedonism, 37�42, 69�70, 210

rational intuitionism, 134�41, 165, 166

rationalism, 77�78, 79, 174, 175, 179,

223, 227

reason, 63, 71�73, 76, 82�81, 121�22,

157�60, 166, 171�74, 179, 181, 196,

212, 222�25

economic, 38

reasons for action

moral, 17�19

personal, 18

Republic

core problem of, 4�7, 12�14, 16, 17�18,

20�23, 73�77, 81, 92, 208

responsibility for self, 182

Ring of Gyges, 6�7, 55, 58

rule of law, 127�28, 170

rules

institutional, 108�11

strategic, 108�11, 119

Sartre, Jean-Paul, 181, 186, 187, 188, 189�92

self-determination, 180�81, 196

self-discipline (self-command), 50�51

self-evidence of moral standards, 134�40

self-interest, 33�36, 45

Sheinwold, Alfred, 110

Sidgwick, Henry, 94n

Socrates, 4�7, 58

soul, see personality

Spinoza, Benedict de, 25

spirit, 63, 71, 72�73, 80

stoics, 123, 202

suicide, 204�05

summum bonum, see highest good

theological voluntarism, 130�31, 132, 133,

140, 155

Thrasymachus, 4�7, 27, 53�55, 58, 99

trust, 47�48

utilitarianism, 93�108, 110, 188�89

act utilitarianism, 111, 113�14, 115�17,

118�22

rule utilitarianism, 111�15, 132�34

valuations, 198

values

absolute v. relative, 167�68, 176�79

creation of, 186, 191�94

incommensurability of, 159�60, 178

virtues

in Aristotle, 83�87

in Plato, 71�73

of a cooperator, 49�51

well-being, 28�29, 57, 59, 61�62, 65�66, 97,

132, 194

wisdom, 27, 71�72, 80, 86

Index 241

Cambridge Introductions to Philosophy) John Deigh-An Introduction to Ethics-Cambridge University Press (2010)

Documents

subject of ethics

existentialist ethics

problems of ethics

psychological egoism

cambridge cb2

central questions of

kants moral theory

moral psychology